As a relatively new participant (long time user) of Math.SE I must confess that my knowledge of what is considered "appropriately non-minor" as an edit is certainly lacking. This question derives from a rejected edit on a recent post where (for the first time) I suggested an edit to a tag - namely adding the number theory tag to a question which is (in my opinion) clearly number theoretic (asking to find rational points on a variety). Admittedly I should have been more clear in my description that this question was number-theoretic even if the OP did not realise (side-point, is this appropriate).
Of the two reasons for rejection, one was
This edit introduces tags that do not help to define the topic of the question. Tags should help to describe what the question is about, not just what it contains.
to which I can only respectfully disagree with the user. The content of any answer to this question will necessarily contain number theory (I now realise this is the pertinent point - see the below edit and comments).
Tags are what leads us to questions we are interested in, so I was of the belief that a tag-modification can in no sense be minor. Moreover I was of the opinion that my change would direct the question to interested parties.
Thus my question is really, is the problem with this edit that it is cluttering the review queues? Is it the case that one should merely wait until they do not need their edits approved when they should edit tags?
Possibly related but unanswered and not identical: Tag edits considered too minor? The comment of dfeuer is nearly pertinent, although in my case I do believe that the tag was not already represented (although I can accept that I simply may not understand the etiquette).
Edit: It turns out that the rejections were about the usage of the NT tag. To clarify, an example of a question that would enter my feed unchanged as NT is If $z | 5x+11y$ and $z | 6x+13y$, then $z | y$. .
Perhaps this is mis-taggeed also, but the question I tagged refers to (as I say in the comments) finding rational points on a curve defined over $\mathbb{Q}(\sqrt{2})$ which is far less "non-advanced" IMHO.
Although at this point maybe this is becoming semantics about what users deem to be "number-theoretic", I believe that the question I re-tagged is much likely to spark an answer with interesting number theoretic content (since we are asking for rational points on some genus $0$ curve defined over $\mathbb{Q}(\sqrt{2})$).