You get ERR: NESTED LIBS.

Same error. I'm guessing Quigibo thought about this before implementing the inclusion of external source programs.

It fails instantly. I wouldn't worry about any bugs regarding this, Quigibo seems to have covered all the bases well.

Not unless you knew what you were doing. You'd have to use ASM.

Axe is actually decently safe (as in RAM clears).

Font editor or token editor? Those are two different things

I'd like to know about both.

Whoops, meant this: http://wikiti.brandonw.net/index.php?title=83Plus:Hooks:9B9C

And no, these are all much more advanced than Axe provides. For the second you need Asm( at least to start up the hook (Asm(EFE44F)). The first would have to be done nearly entirely with Asm(.

So a rundown:

For the first option, you're modifying the OS to replace the built-in font with a new one. That means you'll have to copy the entire page where the font is stored (I don't know which one that is) to a swap, clear that page, then change it back to the way it was (but with the new font). None of this you can do with Axe (except with Asm(), and since the location of the font could change, you'll have to release a separate version of the patch for each OS version/calc model that differs, including future ones (unless you're planning to let some users brick their calc).

For the second option, you're creating an OS hook that basically intercepts whenever the OS displays a character. You'd need to store your new font somewhere constant (which is why this program would have to be compiled as an application), then write a routine (mostly with Asm() for the actual hook, then copy that code somewhere and finally run Asm(EFE44F) to start up the hook. This is a lot safer than the first option and it can't go wrong (since the font hook is an "official" (documented) hook) unless TI screwed something up in an OS update. (Granted, that could happen.)