That page makes it sound like the text in images isn't indexed. I wonder if it's just a matter of priorities, or if there is some other reason for not indexing text in images?

Both macOS and Windows have very good text recognition APIs now. Indexing text in images is starting to feel like table stakes for notetaking apps.

Last time I checked nvalt was working just fine.

Brett has been talking about nvultra for many years...at the time of creation nv and nvalt were novel, but I believe they've been superseded by things like obsidian and notation velocity. Still, best of luck of nvultra.