> I'm not going to stipulate that those giant hooks ARE descenders
You don't have to. I am telling you that these codes were assigned to the extended pe variant that occurs in Abkhazian, and have always been so annotated.
The extended pe variant that I have seen in Abkhazian has the giant hook on it. That is a different diacritic deformation than the descender. If a pe with descender exists, it is not the same as a pe with the giant hook. Those are two different LETTERS. This (letter identity) is a different thing from "what is used in Abkhaz orthography".
> We have seen examples where hook and descender are distinct
If you have examples of two distinguished pe variants occurring within the same Abkhazian text, then there is a problem.
I don't know whether PE WITH DESCENDER needs to be encoded or not, but it is certainly not a glyph variant of PE WITH HOOK. That is the kind of overunification which I will never consider to be acceptable.
--
Michael Everson * * Everson Typography * * http://www.evertype.com
