Should there be an apostrophe in the name?

This changeset edits a way around the tennis court so that court is now duplicated and the boundary fence is unmapped. I've reverted this (and added your capacity tag to the other way instead).

Did you look at the surrounding map data before making this change? I'm pretty sure this is basically a flat toll plaza with no bridges; it looks like the original mapper decided to link all border post exits with all toll booths (on this side). While the original mapping is incorrect, your edit here just makes it worse.

This change looks like a mechanical edit (in that you are changing lots of things systematically without necessarily properly reviewing each change). There is now a page on the wiki about these (osm.wiki/Automated_edits). Unfortunately I can't easily check whether you've made other errors like this, or whether this was just a rare mistake.

I don't know, I've never used iD. JOSM explicitly asks which portion of the way should retain the history when splitting an existing way. If iD doesn't allow that, then it might be better to change the way you use iD to avoid this issue, or use a different editor.

Perhaps there's no explicit restriction, but the only thing they lead to is a road on which pedestrians aren't allowed. I think there are lots of similar maintenance access steps on motorway/trunk road embankments and cuttings, and I couldn't spot any others that were explicitly mapped. (One example is some existing steps next to the bridge over the A14 in Girton.)

You've manage to complicate the way history here by reallocating the way originally representing to the cycleway over the bridge to instead represent the cycleway adjacent to the bridge, and creating a new way for the bridge. Can you try to avoid this in future? I presume you did this by extending the existing way and the splitting it, with the history being retained on the wrong segment after the split; you should check that history goes to the correct section when splitting ways to avoid this sort of issue.

I don't think these are for normal public use, so should presumably be tagged "access=no". I tried to find existing examples to compare to, but it looks like this sort of access stairs haven't generally been mapped anyway.

This changeset doesn't actually contain the changes it claims to, because the intended changes almost entirely replicate those in 83547340, which was uploaded 5 minutes earlier (sorry, I beat you to it :) ).
The changes that actually remain are modifications to a stream (marking part of it as in a culvert and realigning it). I'd recommend in future that when you split ways, you try to ensure the main part of the way retains the old id. In this changeset I was left wondering for a bit why you had created a new stream to Longstanton, until I realised that the new way replaced an existing one which you had shortened and almost entirely realigned.

If you're changing the "highway=construction" tags because it's open, then you should also remove the "construction=cycleway" tags. Also, the "oneway=no" tag you added to one of the ways is redundant.

I think the tagging in use prior to this changeset was probably more accurate (and the description at highway=track agrees). I guess that you made this change because the rendering looked funny with a track leading to a driveway, but the correct way to fix that is to change the tagging on the other part of the driveway to highway=service, service=driveway (with the existing foot=yes indicating public access on foot).

There are a couple of issues with this changeset. Firstly, you've removed connections between the path and the road at the roundabout. This means that routers will no longer be able to use the paths for routing, as it looks like it isn't possible to rejoin the road (or a pavement adjoining the road) at that point).
Secondly, I cycled past here a couple of weeks ago, and at the time there was no accessible footpath on the west side of Robinson Way because it was part of a construction site. Has that now changed? If not, then I don't think a currently non-existent path should be mapped through the construction site.

The house numbers were already present on the building ways, so this changeset has added duplicate numbers. It seems that separate nodes were only being used here where one building contains multiple addresses. I've reverted this change in 83081728.

The paths added in this changeset should have been marked as private. I've fixed this myself, but I haven't added any notes detailing these restrictions (such notes are currently present on other footways in the college).
It also adds a tree in the middle of the river. I don't recall there being a tree in the river here, so it seems likely that either the tree has been mapped in the wrong location, or (less likely) the river is inaccurately mapped. Can this be checked/fixed please?

Also, I think you've mistagged the patio area (or whatever it is). The access tag is on the outer way of the multipolygon, but should probably be on the multipolygon itself.

This changeset has been reverted by sladen in 82308605. If you haven't discussed this data with them, then I suggest you start that discussion.

I wouldn't describe smb1001's changeset (75018301) as "armchair mapping damage" - that implies to me that the changeset was incorrect for reasons that would be apparent on the ground. The actual disagreement here is over the relevance of the data to OpenStreetMap, including whether it's relevant to include the name a mapper gave to their own (or at least someone's - I don't know that it's yours) private back-garden BBQ.

Personally I wouldn't include the name tag, particularly as it currently prevents the house number appearing in some renderings. I'm also not sure that this is really the intended use of the 'amenity' tag. In any case, since you clearly disagree with smb1001 about what is appropriate to map, then you ought to have some sort of discussion with them, rather than just reverting their edit. (Perhaps you have had that discussion - if so, then it's worth mentioning so that other people know it's happened.)

You forgot to adjust the bus route relations; I've updated these to use the new road now.

In this changeset you enlarged the roundabout but forgot to adjust the layout of the service road or under-construction layout cycleway to match. I've changed these to match your road alignment (without realising at the time that the road realignment was a very recent edit).

I don't think this is a correct fix - you've just changed one incomplete mapping of the car park into a different less accurate mapping of the car park (the 'dead end' you removed was the exit from the bottom of the spiral ramp; you've redirected it to connect to an exit from the ground floor).

Can you confirm whether or not there is pedestrian or cycle access between these two roads? Your edit implies that there is no access at all, but it seems unlikely that there wouldn't even be a gap for pedestrians.