Currently states only english language is case-insensitive.

When in fact

The text search subsystem treats "A" and "a" as the same character (in particular, for text search queries and for text-indexed documents). The same relationship applies for all characters in [A-Z] <=> [a-z].

Aside from those characters, all other characters are considered unique / not considered to have a case-equivalent. Improving this is tracked at SERVER-8423.

So, you're right, it's not language-specific.