To me it feels obvious — making them the same codepoint makes case conversions require knowing which language the string is in. Making them separate codepoints does not. The only important question is whether Turks use separate keyboard layouts for typing in Turkish and English, because if they don't, this does also make things complicated, but differently.