There are a lot of questions on this site which ask for evidence. I think there's a substantial difference between questions which ask "Is X true/likely?" and "What is the evidence for X?" This is a site for experts, so I think it's great that we have so many evidence questions, and so many people who can answer them!
How should these evidence questions be answered?
Obviously if you can provide evidence that confirms the idea then that would be great.
What if you can't provide evidence? My natural intuition is that if you begin your answer with "There is no evidence for that. But I think ..." then that is an acceptable answer, and that you can continue on with your own theory. You may even be wrong about what the state of evidence is, but it's still an acceptable answer.
I don't think that it is good to give your speculations to an evidence question without explicitly stating what you believe the state of evidence is. In an evidence question that would be a bad answer. Evidence-less speculation doesn't add value to this site, and there are lots of other places on the internet where that kind of speculation can thrive instead. Would it also be a non-answer?