https://bugzilla.redhat.com/show_bug.cgi?id=2262410
--- Comment #27 from Oleg Oshmyan chortos@inbox.lv --- Why do you keep mentioning English? libass is looking for a font to display numbers in. There's no English involved.
How about simply taking a look at "lang" for a font?
It's empty. You've removed it.
"lang" object indicates what character coverage is needed to represent for. "en" almost covers Latin characters loosely. If you know what characters represent a language, you don't need to check it with FC_CHARSET one by one.
libass still needs to look up every glyph. Language coverage has some correlation with glyph coverage but one does not guarantee the other. Font designers can't draw every single glyph; Unicode keeps getting new code points; languages themselves are loosely defined (are digits "English"?); fonts can be subset; etc.