Genuinely curious: Are non-ascii characters also case-insensitive. With Unicode comes different case-sensitivity rules according to Unicode version and locale.
I strongly suspect the systems that are uppercasing everything were not written to handle unicode in the first place though.