The zero comparative marker and the non-zero one should be more or less interchangeable. (The etymology of the non-zero marker doesn't matter.)
(A message asking to list such languages was originally posted by Lisa Bylinina; check under number 2, it's written in Russian.)
This question appeared from the exploration of the relation between low degree markers and the comparative forms cross-linguistically (see also Comparative markers coming from low degree markers ("attenuatives")? (List such languages.)) and the observation of an example--the Tatar language--where indeed the low degree marker is the source of one of the comparative forms:
Tatar is a parallel case [to Bulgarian] I discovered this summer while doing fieldwork on Tatar. not absolutely parallel though, because this low degree marker is optional in most of the comparative environments, but seems to carry no low degree meaning anymore
which is also an example exhibiting the kind of situation queried about in this question:
[This Tatar marker (coming from a low degree marker)] is a non-ambiguous variant of zero comparative morpheme (which is ambiguous with the positive form)...