https://wordfinderx.com/blog/languages-ranked-by-letters-in-...
I have no clue as to the accuracy of this website, but accuracy isn't something we strive for when making ridiculous comments on the interwebs, is it?
Japanese could be combined with the Hindi character set to yield base 96, which is fairly convenient. Cyrillic would be harder -- perhaps the best options there would be to drop a character to yield base 32, or perhaps 3 characters to yield base 30.
I'd argue that base 60 is probably the optimal number base for nearly any use (with base 16 or 64 as close second and third for working with binary data). Hindi's 50 characters combined with our 10 Arabic numerals could indeed be a great way to get there.
> That’s one hundred billion supernovae per century, or a billion per year, or about 30 per second.
7 characters of base26 gives you 8 billion combinations. "aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa" requires ~1e48 events when going by the actual non base26 scheme. So I wouldn't be worried.