Adjust static font details so that we can properly display Arabic
contextual letter forms. And so that alphabetical ligatures are loaded
from language-specific fonts.

This is mostly needed because I didn't realize that font table OS/2 coverage for Arabic includes three separate coverage bits. This means we currently won't show some glyphs that are in the Arabic font that are not also in the base font, like "ﷸ" or "ﻼ"

These changes have no immediate benefit. Arabic Presentation Forms-A are generally just used for Arabic words that have been "shaped" (contextual form changes and ligatures) and we don't do that yet. Arabic Presentation Forms-B is similarly post-shaping diacritics, and contextual letter forms.

Removal of "Alphabetic Presentation Forms" coverage bit from Armenian is just because that block contains standard ligatures for Latin, Armenian, and Hebrew. They are also in the Latin and Hebrew fonts and I'd rather they be taken from the specific font, not always from the Armenian one.

Note this is only very small data changes to a struct that I created very recently, so it is possible that I will commit this without review.