I just saw this meta post concerning a tag for NFU or more generally New Foundations. There is good reason not to have a tag for too specific tags, but then I also noticed that there is a tag for univalent-foundations. At present it has only 7 questions asked by only 5 people, so I think there is no reason to keep it.
I think the reasons given to potentially remove the univalent-foundations tag are misguided.
There is good reason not to have a tag for too specific [concepts]
I would argue that in some cases there are very good reasons to have tags for specific concepts. In fact, I believe that at times these are better tags than many of our broader ones. Certain specific mathematical constructs/objects/etc. come up again and again in questions on this site, and it is often useful to group these together under a tag. Sometimes they go by several different names and notations in the literature, and so relying on site-search to find relevant questions is not always optimal.
A couple that I have been involved in the creation of are the-baire-space and sorgenfrey-line. I argue that these are much better tags than the much larger compactness and connectedness, which have great ambiguity to their usage. And our extremely broad tags (calculus, analysis, topology, number-theory, etc.) only serve to delineate the various branches of mathematics, but often say very little about the content of a specific question.
At present it has only 7 questions asked by only 5 people, so I think there is no reason to keep it.
The size of a tag really has no bearing on its usefulness. Tags "are for sorting your question into specific, well-defined categories" in order to "connect experts with questions they will be able to answer" (see this Meta Stack Exchange answer). That a tag is small only implies that (currently) few questions have been asked under that tag. It does not mean that categorising these questions together is without merit.
That said, I am personally uncertain about the usefulness of a univalent-foundations tag. My concern is that it is not clear to me how this tag differs from the homotopy-type-theory tag. I am not an expert in these matters, but my understanding is that there is considerable — if not nearly complete — overlap between these topics. So from my standpoint the question is whether univalent-foundations should stand alone, or be synonymized with homotopy-type-theory (with one or the other as master).
But since univalent foundations (HoTT) constitute a specific, and perhaps currently non-standard, approach to mathematical foundations, using a separate tag to separate these questions from those about the currently-more-standard set-theoretic/ZFC foundations seems appropriate.