Which polygon? There's a few in the change.
Certainly there was some Merkaartor shenanigans involving an unreferenced way. I'll look at fixing it up manually.

For anybody wishing to check, https://en.wikipedia.org/wiki/List_of_Outer_Hebrides says, "There are other outliers with cultural links to the Outer Hebrides that are not part of the archipelago itself. These include the St Kilda group, which are quite distinct geologically and no longer inhabited, Sula Sgeir and North Rona to the north and isolated Rockall, which is 367 kilometres (228 mi) to the west of North Uist."

Sorry - it looked like you were deliberately ignoring the fix. I was clearly in a bad mood that day!
As an aside, do you agree with me that Rockall and North Rona/Sula Sgeir should be removed from the "Hebrides" relation, whether or not they are administered by Comhairle nan Eilean Siar? I've never considered these outliers to be Hebridean, and Wikipedia concurs.

Please stop breaking the continuity of North Uist. I've repaired this twice in two days; I don't want to keep re-doing that.

What are you trying to achieve here? I don't understand why there's a big circular rough outside the bounds of the course, nor why its inner way isn't actually within it.

"Plait Lodge"? I've always known it as Pait Lodge.

I think that would be from OS 6" series around 1900 or so. It's likely that some of these are no longer visible from above, but there may still be evidence of their routes.

Construction speed limit of 40mph is still in place.

Why did you change this name? It was correct to begin with.

Thanks for the correction here!
I think I'm finished for today, and I will be offline Saturday and Sunday, so give yourself "free rein"!

This multipolygon was broken by having relations as members instead of ways, which is not defined for multipolygon - I've also corrected a similar error in other recent edits. Please check OSM Inspector (or another tool) when editing relations. Thanks!

Thanks - fixed.

Thanks - that's great, and it makes OSM Inspector (Areas) much happier!

What's the purpose of relation 15475169, which has only a single "inner" member that duplicates the tags of the relation? Something is clearly wrong there, but I'm not sure of the appropriate corrective action (and aerial photos don't clarify it for me).

This change has seriously broken the heath area relation (osm.org/relation/13417952). Can you fix that?

Yes, that's exactly what I mean about ways that start off identical and then diverge. I've not used an editor that adds to multiple ways simultaneously - I didn't know that there are some that can do that. I'm certainly seeing unintended overlaps and gaps between areas. And also linear features that clearly ought to be the delineation between areas (fences, walls, hedges, etc) getting out of sync with their respective areas. I always understood it to be best practice to use these boundary ways to form the areas.
Having multiple ways sharing the same nodes certainly makes it very difficult to select the way you actually want to edit much of the time.

It's a judgement call whether the scrub is a single feature or not - feel free to separate into smaller relations if necessary.

The duplicated edges tend to fall out of sync when edited so that abutting areas end up overlapping or not quite meeting. This particular area was highlighted by OSM Inspector due to touching rings, so I fixed up some of its duplicated boundaries whilest editing.

Here's how I fixed the ones you introduced here:
osm.org/changeset/130383971
The "inner" ways were present, but the relations were missing the "outer" way (the loch itself).

E.g. see the history of Loch Shiabhat:
osm.org/relation/15013045/history

Ah, great. :)