it was already decided (I did not change that, there was already a polygon about it) that they would be isolated at admin_level 2 (even if they are not isolated as recognized states); this is also true for other condominiums (e.g. between Luxembourg and Germany, or between France and Spain), and in other disputed areas (e.g. between China and India or Pakistan and India, or Sudan and South Sudan, or Sudan and Egypt);
Note that the recognized countries can overlap in these areas, according to the wiki This is here an unsolved case with not even a clear "de facto" state, just peaceful cooperation between the two countries in this area.
There are other disputed areas in the world, msot of them maritime on the EEZ or the continental plate extensions.

It's a fact that the wikipedias links were incorrect (some with old or broken redirects)

I did not modify or added any border or changed the existing classification. These are common edits that any one does in OSM.

@Someonelse, check the datetimes, everything on the talk-es list came hours after that changeset; I've replied to them in due time and acted accordingly...

@SomeoneElse: you posted that question after. And I replied now to the list; I had no message from the list before that changeset.

I've replied to them. Or did not see any question directly asked to me.
Was there any post in the list with these "questions" I'm supposed to reply?

Which question have I not replied ?
Also I've only changed the enclave which was at admin_level 7 but also over another comarca at level 7 containing it (comarca de Ebro), and not found any reference that cites it as a comarca. It may be a only subcomarca, or a traditional comarca, but not with an admin status today.
I've not removed its border, it's still there

You semm to have never seen that these were NOT the same type of relation.
You missed the distinctive tag which was not "boundary=administrative" but "boundary=administrative_fraction";

"type=boundary+admin_level=*" alone is not enough to distinguish relations (see for example the religious boundaries which also have their admin_level...)

the "_fraction" type is different; applciations that just want to see full comarcas will ,not load the fractions which are separate kind of objects, not adminsitrative, but related to correctly represent the relation and partial containment of levels 6 and 7 in Spain, where it is not exact.

These relatiosn had differnt contents, coverd distinctive areas. The names indicated there are just indicative, I could have put none at all, but it's better for editing users trying to read it, and that's why these names had disambiguation suffixes (nont intended ot be rendered on any map, and not rendered at all in fact. But still useful in statistical reports. This just proves that I was really aware that provinces do not contain full comarcas, this is alsmot the case but not everywhere. There was no error at all, these are "building blocks" useful for data checkin, check unexpected overlaps or incompleteness, and also make relations easier to interpret while still permitting a containment: each "fraction" have two parents not just one, and not the same two parents (they stilll form a graph as a tree with common branches and a common ancestor, but the graph remains acyclic).

attention l'ajout de ce parc ^naturel prsè de Tournais a cassé pas mal de relations sur la frontière franco-belge, à la fois en France et en Belgique, j'ai réparé un endroit de frontière franco-belge coupée en deux concernant la France sans reporter les morceaux dans les relations parentes, mais je n'ai pas vérifié tout en Belgique ou au Luxembourg si c'est cassé aussi...
Avant de scinder un chemin dans JOSM, ou d'en fusionner deux, surtout une frontière internationale ou régionale (où les dégâts sont majeurs et impactent pas mal de chose), il faut toujours charger les relations dépendantes (CTRL+ALT+D) pour que la scission fonctionne correctement dans toutes les relations.
Pour la fusion de chemin il ne faut pas fusionner si les deux chemins ne sont pas tous dans les mêmes relations (JOSM affiche un "X" dans la liste de numéros de membres si une relation n'utilise pas un des chemins quand on les sélectionne tous les deux) et qu'ils ont des tags compatibles (aucun en conflit)

Dans iD on n'a pas ce problème: la scission marche en principe toujours car toutes les relations dépendantes sont chargées quand on scinde un chemin, sinon pour la fusion il la refuse si les tags ne sont pas compatibles ou si les deux chemins ne sont pas membres ensemble des mêmes relations.

I'm not "forcing" changes, I'm adding what is missing. I still need some tags for these, and have just used what is currently available, not changing them.
But I do agree that there are several comarcal definitions and I have never contested that. This does not mean that they cannot be added at all and distinguished. Only one comarcal division is being made for now (where there is none in OSM and they are clearly missing). We can add more and it's not at all a problem for me and should not be a problem for you. Those are already used, referenced in other collabarative sites already linked by OSM (notably Wikimedia, Commons and Wikidata where there's also lot of work to sort these as all is mixed and confused).
I'm not expecting to work alone. But there's no contestation that I do not invent anything, and that I am just filling missing gaps of information, not destroying anything.
It's quite simpelr to change some tags if we need several definitions, it's a matter of a few seconds; but restoring or rebuilding completely from scratch is a severe loss of time for everyone. The work done or to do require lot of time, desctroying it gives only negative value.

I know it is difficult (and the categories used in Wikipedia or Commons or in Wikidata are not making the distinction easy, as all comarcal types are mixed, except in some regions).
This is lot of work to do in several incremental steps to sort all these correctly. This changeset was one of the incremental step needed to help building these distinctions.
The next step will be to map the mancommunidad (cooperations of municipalitiies) and this is exactly similar to French intercommunalities, tagged using "boundary=local_authority, plus some distinctive "admin_type=*" for them).
Spain is not more complex than France, it just requires clarification and extensive work to sort all this.
But having only regions, provinces in Spain to group municipalities is clearly insufficient.

Note that even if there are multiple comarcal delimitations in regions that still don't have their own official ione, and still use traditional comarcas (still used) and "touristic" comarcas promoted by the deputation of the provincial junta, all of them can coexist in OSM; this is just a matter of distinctive tags.
For those qualified as "boundary=administrative", this should be the regional definition according to the regional law if it exists. Others can be tagged as "boundary=traditional" or similar (with a specific case for older comarcas that are actually subcomarcas of more recent or larger traditional comarcas.
provincial/touristic comarcas may also have their distinctive tag like "boundary=touristic".
In all these case we can also add a more explicit tag like "admin_type=*" (to be chosen by the Spanish community).

And a "description=*" filed to exhibit this in human language (possibly translated in several languages, including regional languages spoken in Spain, notably Galician, Asturian, Extremaduran, Aragonese, and Catalan with a relevant language suffix like "description:ca=*")
I do not see that as a big problem.
some comarcas may have identical definitions, from the region or the province and in that case they still qualify as administrative. The "administrative" level 7 should still be complete, even if for now it is fixed in some regions by the agrarian definition (which is also administrative but for a limited purpose but still commonly used as long as the region does not officialize its own delimitation).
Other comarcal delimitations dintinguished from the official regional one will no longer use the "administrative" type, but a different type (boundary=traditional is not convenient, we could have boundary=agrarian for them).

The technical problem of the OSM list server is in their DNS, and the fact the MLM breaks DKIM, and DMARC, leaving only SPF. Not all mail providers support SPF (note: Wanadoo is NOT historic, it is a brand of Orange, a major ISP in France, and lot of French users are forcibly unsubscribed without notice as the mails sent by the OSM MLM are bouncing systematically. Technically the digital signatures for DKIM and DMARC are broken, so the recipient service provider try to check if these mails were usurpated or modified. As all supported check fail because of incorrect installation of OSM MLM servers, the talking tools are not working as they should by mail, there are also many other places and OSM has still not worked seriously on consolidating them; tons of OSM users get then blocked unfairly even if they had some talks, they can't reach everyone they'd like to join).
Only this channel (changeset talks) seems stable for now, as well as the private messages on the OSM account of this site. But they are interpersonal tools, not appropriate for mass delivery to all concerned people.

Also you talked with me positively, so after thinking and searching I started to do the job (I was instructed to do that by someone that told me he had ionformed the Spanish community in messages in Spanish).
I received no message at all, I was just blindly reverted without notice, and even reported in an attempt to block me, even if I had made absolutely no damage at all.

Note: I have no problem for reading Spanish, I just avoid writing talks in Spanish. And not all spanish users even talk Spanish as the country is highly multilingual, including in Aragon (I would have much more problems reading Catalan or Basque, I may interpret Aragonese incorrectly). To limit errors when I write something I just use English but you can reply in Spanish if you prefer, it's not a problem. I have consulted only the talk-es archives but cannot subscribe it.

Also the difficulty of the task is not a problem for me. Of course it requires patience and lot of work. I can do that, patiently, calmly.
There may be some intermediate errors but that's not worse than when there's nothing before. Yes the boundaries in Spain are very complex, I do not challenge that. It does not mean they cannot be represented. I do not destroy anything. I'm not spamming arebitrary data, all is checked many times by various methods and lot of searches.
And I've tried to join the talk-es list, this failed, I never received any message (because of installation bugs in the OSM maliling list manager that causes the MLM to receive lot of bounces: the OSM servers are not properly installed and they forcibly unsubscribe many people)

And may you don't know but I have the legal document named "DECRETO LEGISLATIVO 2/2006, de 27 de diciembre, del Gobierno de Aragón, por el que se aprueba el texto refundido de la Ley de Delimitación Comarcal de Aragón", published on 30 December 2006 in BOA n° 149. I use it to find decrepencies (e.g. missing "alt names" used by the region, different from those used by IDEE, and forgotten in OSM).
There are various municipalities that don't have their admin_center node linked which I fix too (this really helps creating other divisions and locating items on the map data). I add the missing wikidata items (or create them as needed, making also links in Wikidata for some relevant wikis and adding the missing classification or fixing a few declarations, already signaled in Wikidata as they reference the incorrect object not of the correct datatype). I consolidate all that together, and mlake sure that all relevant sources are linked together and easily verifiable.

No I do not ignore that. I spoke about the agrarian ones but not for Aragon.
I use the same sources that the Spanish community has developed in OSM, in Wikidata, in es.Wikipedia, I look at them, search the official sites, look for legal texts. These are thousands of clicks and hours of searches.
And I was really talking with other Spanish users (the DWG can even consult my talk page here on the OSM site)
I've not "invented" new borders, just created a few separate objects to manage the creation of missing borders and correct classification of municipalities. This is a step to check them (even the existing comarcas are not complete everywhere as you think).
Now the minor name error you talk is not even in the boundaries that you have kept, only in the temporary subjects (that are clearly not "duplicates" and were tagged specifically and differently from comarcas, and not described as comarcas, with additional note to exhibit it more clearly, but you ignore that).

So in conclusion I do not ignore the community, I take their input in consideration; beside the minor name incorrect in the subrelation that could have been fixed, but was blindly removed with everything else, without any prior talk to suggest easy fixes and explain what we can do...
I have various messages in my talk page here about these, they were positive and did not go to an unexplained blind revert (which also restored many errors I had fixed) but only to some fixes.

This was already fixed in that subrelation.
And no, other Spanish users contacted me to say thanks.
I was asked to compelte the work in Aragon, and this revert just complicated the task, These subrelations were needed at least as an intermediate step to fix various things, including dixing many names, missing admin_centers, adding references (wikidata and so on), and debreaking some broken boundaries (I checked them in the Spanish cadasdtre to see if the conflations were correct as these limits are still approximative and deviate frequently between two neighbouring municipalities, sometimes creating articficial holes.
The coverage has grown. And I use the msot relevant sources agreed on several Spanish communities (OSM wiki, Spanish Wikipedia, and Wm Commons, checking the dates as they have changed over time or because some of them are different proposals of comarcalisation or because they are traditional and not endorsed by the agriculture ministry, or the juntas and deputations of provinces or regions which use different definitions; I was told to use iun preference the agrarian comarcas when there are conflicts, instead of touristic comarcas proposed/promoted by provinces, differently from the regional comarcas and also differently from multiple layers of traditional comarcas, various ones being in fact subcomarcas of another traditional but newer and larger comarca).