Personally, I find linking to some source, be it wikipedia or anything else, to be helpful in many cases. In particular, it isn't always clear to me whether the questioner is using the terminology in the same way as I am. It isn't always even known to the questioner that the term may have multiple meanings. I'm sure one can come up with extreme examples of this, but I won't bother here.
Linking eliminates the need to go through long strings of definitions just to make sure everyone knows what you are talking about. However, I do agree that it isn't necessary if the term is well-known. It can be overdone (I decided to remove wiki links in my question since the terms were unambiguous), but I don't think it's really much harm when it is done because people who stumble onto the question without the background can still get something out of it.
I have to disagree with your statement that we should avoid posting any links to Wikipedia because they might be incorrect. It seems to me that if there are errors, it is worthwhile to the larger physics community as a whole that these get pointed out and corrected, and a link from here is probably fairly likely to do so. I don't have much comment on your other issues with Wikipedia, as I haven't experienced them enough to have an opinion.
However, I agree that linking to a review paper, at least when available, is a better idea. Ideally Wikipedia would be the more error-proof of the two, since errors can be corrected immediately rather than only via an errata, but practically this rarely happens. With that in mind, I find myself ultimately supporting this proposal.
This post has been migrated from (A51.SE)