Why not just allow a question to have both a more specific tag and a broader "umbrella" tag--
Examples--
Both "class-e-airspace" and "airspace"
Both "stability" and (some broader tag-- seems like we're still trying to decide which one to keep and which one to "burninate"-- either flight-dynamics, or flight-mechanics)
Both "sailplane" and "glider" (although arguably, it would probably be fine to get rid of the "sailplane" tag, as most people searching for questions pertaining to sailplanes might be interested in almost everything that comes up under the "glider" tag, since we don't tend to get too many questions about cargo gliders, paragliders, hang gliders, or the space shuttle). Removing the "glider" tag from everything that could be legitimately given the "sailplane" tag, on the other hand, would be a really bad idea, as it would transform the "glider" tag essentially into a tag functioning to cover "all gliders except sailplanes".
The idea of allowing a question to have both a broader "umbrella" tag, and a more specific tag, doesn't seem to be against the ASE help center guidance as I'm understanding it-- I see the caution against "meta-tags", but looking at the specific examples given for "meta-tags", which serve essentially no useful purpose, that doesn't seem to be the same to me as these somewhat broad, but still useful, "umbrella" tags I've used in the examples above. Either the "umbrella" tag or the more specific tag could stand alone as the only tag on any question that they are correctly attached to, and thus neither are "meta-tags" in the sense described in the ASE help center.
This idea that tags should not be subsets of other tags is coming as a surprise to me. I don't support it--
Sure, some tags have arguably gotten too broad (e.g. aircraft-physics?). But the idea that we need to "do something" about all cases where a question carries two tags, one of which is clearly a subset of the other, seems the wrong approach. Both tags often potentially play a useful role to someone searching through the question database.
Another example has just occurred minutes ago. An edit was made to delete the tag "instrument-procedures" from this question When can I proceed inbound in a hold in lieu of a procedure turn when cleared for the approach? . How is this not an example of a question about instrument procedures? The idea, apparently, is that it should not bear the "instrument-procedure" tag because it has also been given the "iaps" tag which references instrument approach procedures? This sort of editing isn't doing anything to make the "instrument-procedures" tag less vague or more specific, it's only making it occur less frequently, which is not the same thing. In fact, in parallel to my comments about sailplanes/gliders above, it transforms the "instrument-procedures" tag into a tag that functions to mean "instrument procedures, except for those pertaining to approaches" -- this is not helpful.
If the community consensus is that this sort of thing is the right way to go, then we need to really think twice about creating new, more specific tags, because of their huge potential to degrade the usefulness of established tags. But it doesn't need to be that way, if we allow a question to have both a more specific tag and a broader tag.
And yes, I've now seen the SE meta post against the idea of hierarchical "trees". It still seems to serve a useful purpose on ASE to allow tags that are clearly subsets of other tags.
The idea that the application of a newly-created, more specific tag to an existing question should mean that it's fine and proper to delete any more general tag already applied to the question, seems incompatible with a system that allows any user w/ 300 rep points to create a new tag. The potential for chaos seems immense.
