I don't think a revision tag will help -- at least, not enough to merit the creation of a meta-tag.
It is laudable to strive for "some accessible guidance on how people coming to the site for help with revision could best find what they are looking for".
It is similarly laudable to strive towards people including all relevant information on questions they ask (particularly, their own work). It is reasonable to expect some effort from newcomers in making them blend in well with accepted practice.
However, in spite of many contributors' extended time investment in these latter points, there is only a limited amount of success. All too often, no response to their suggestions is recorded. (This eventually leads to frustration, unfriendly comments and meta fireworks about PSQs (problem statement questions, consisting of only a problem statement).)
I thus have no reason to believe that a revision tag will somehow tread a different trail compared to these other suggestions of similar positive intent. I think it will just hit the wall of internet reality (a nice oxymoron, by the way). That's sad, but it's also true. We won't win anything by being oblivious to it.
I wish there were a more positive story to tell, but I'm afraid this is it. Any proposed resolution of the above problems will have to start right at the bottom with a well thought out contemplation of what Mathematics.SE is, and even then the conclusion may be that there is nothing to do about it.
Addendum as the comments by Tom Oldfield suggest that my point may not have been completely clear.
I have interpreted the intent of the revision tag as being two-fold:
- Firstly, and primarily, it should help find others revising for their exams to find the issues they are working with;
- Secondly, it should help answerers provide a clue of what type of answer is requested.
Since MSE is not, and does not want to be, a course reference, we are to expect that these questions in fact aren't intrinsically very different from ordinary questions ("What is the definition of X", "How to apply Theorem Y in case X", etc.) except maybe that they are more appreciative of universal solutions.
But since there isn't a lot separating normal questions (which regularly get general answers too) from revision questions, it may actually have an adverse effect on revising students finding the information they seek (misconceptions along the lines of "I'm revising, so my answer will be tagged with revision." lurk).
To have a meta-tag that is bound to be very localized, especially when compared to the prospected universality of the answers, seems to be superfluous. Since the first point isn't properly served by it, and the second can be adequately dealt with in the body of the question (as user79365 suggests). The fact that the questions and answers will be generally of higher quality can be reflected by the vote counts -- and it would be wrong to have a "revision" tag when its best purpose would be to indicate a prospected higher quality, which is a different matter altogether.
In conclusion, the revision tag is of positive intent, but it will not have the desired effect. It is in this aspect that it compares to the other initiatives referenced above.
(Thanks to Tom Oldfield for making me realise the incompleteness of my answer.)