Yes, as the OP suggests, I think #2 could be used when one intends to mean 'amenable to comparison' in a scientific or mathematical context (as in the comment from @Gregor y related to programming). Consequently, no, the two pronunciations cannot be used interchangeably.
I am assuming that the OP spelled out #2 as 'Com-PAIR-ah-bul' intending it to rhyme with 'unbearable', i.e. /kəmˑpɛəɹəbəl/ in IPA. If so, this would be different from any of the pronunciations in @nohat's attempt to translate the question into the two sets of pronunciations found in dictionaries: a) /ˑkɑmprəbəl/, /ˑkɑmpərəbəl/; b) /kəmˑpærəbəl/, /kəmˑpɛrəbəl/.
If the OP did indeed intend #2 to be pronounced as if saying the word 'compare' then adding '-able' as an afterthought, I think it would be an invented pronunciation, but one that would be widely understood without necessarily being recorded in dictionaries (aside: is it called a neologism, when it's a new pronunciation but not a new spelling?).
Dictionaries do give the meaning 'capable of or suitable for comparison', for instance, Merriam Webster gives that meaning with the following example: 'The situations are not at all comparable.'
However, I think there is room for the OP's #2 pronunciation when spoken in scientific contexts, e.g., "The experiment must use standard methodology so that output data is comparable /kəmˑpɛəɹəbəl/". Without this invented pronunciation (rhyming with unbearable) if the speaker used any of the four dictionary pronunciations in @nohat's answer, it would beg the question 'Comparable with what, specifically?'
Even if the speaker stated the object of the comparison, e.g., "...so that output data is comparable with other studies," the audience would think the speaker wanted the results to be similar to other studies, not that they wanted it to just be possible to compare the results with other studies (even if they turned out to be different). So there is a role for the OP's new #2 pronunciation.
To disambiguate without any pronunciation clues (i.e. in writing) it would be advisable to either write 'amenable to comparison' in full, or sometimes context might be sufficient. For instance:
- If there is no object of the comparison, it implies meaning #2.
- If the comparison is set in the future, it tends to imply meaning #2. For instance, "The experiment must use standard methodology so that output data is comparable with future studies". However, setting the comparison in the future could still be ambiguous, for instance, "We hope our output data will be comparable with future studies."