Consider changing `Char` to represent a code point rather than a UTF-16 code unit

Reference: https://github.com/pure-c/purec/issues/34#issuecomment-497961729

Exposing `Char` as UTF-16 code units is problematic for backends that need to use something _other_ than UTF-16 as the underlying representation. Would it be better if `Char` was encoding agnostic and represented a code point? What would that mean for `purescript-strings`?