This has been proposed before: https://forum.inaturalist.org/t/improving-inaturalists-nomenclature-taxonomy/36143
There didn’t really seem to be much buy-in at that time. One of the major concerns is that editing IDs can break the flow of comment threads: for example, if I add an ID to Tritia, you comment ‘No it is Ilyanassa’, and then a taxon swap edits my ID from Tritia to Ilyanassa, your comment no longer makes sense for future users. However, if future users can see that the taxon swap happened after you made your comment, then it will make sense to them. This particular case would not necessarily be all that confusing, but you could imagine more complicated cases like a sequence of multiple taxon changes in a row that would render the comment discussion total nonsense unless it is placed in chronological order.
This is not a problem, it i the system functioning intended; unless the IDer withdraws their ID it should not be eliminated by a taxon change. Sometimes experts add a maverick ‘contradicted’ ID that turns out to be correct!
Are you saying that in the sense of the current taxonomy all Tritia IDs in ~North America are actually Ilyanassa? If so theoretically that could be done as an atlased split, and maybe should have been done that way initially. If that is the case you should flag the genus Tritia.