Sometimes, I come across answers (not comments) that shows a proof that the code of the OP is working fine. Of course, not all such answers are correct, since the difference in version of the interpreter/browser/whatever runs the code may cause the difference in behavior. However, in case that such difference is no way related to the problem, is "it works fine here" considered an answer?
Not really. It doesn't answer the question...
It does makes a great comment, however.
For it to be an answer, you need to detail what could cause the issue, not that you can't reproduce it. You can explain that the different versions are not the issue.
Yes and no.
If in fact the basic premise of the question is flawed, then the correct answer is "actually, nothing is wrong here". It's the only valid answer that can be given, as it is the truth.
Ideally, of course, such an answer would also be able to inform the OP as to how they misinterpreted their results, or perhaps how they could have misconfigured their system to get different results. This would have to be part of the answer in order for the answer to actually get an upvote from me.
That's my view within the scope of a posted answer.
However, the only scenario in which this makes sense is where an OP has posted a localised debugging question. "What's wrong with my code?" or "My code doesn't work". These are not okay on Stack Overflow and, as such, the question should be closed with no answers given at all.
Comments are a poor place to provide proof of working, like console output, mixed code and commentary, etc.
It depends on context. I don't blindly mark them as non-answers, because dense, code-ridden comments are tough to deal with, and such non-answers may have value.