What I have repeatedly experienced in my short time on this Q&A site is that I pose a question, and receive serveral answers, each of them contributing some perspective of value, but none fully answering the question to my satisfaction.
From all the comments, discussion and answers, I eventually end up forming a conclusion that satisfies me overall. It appears useful to take the best parts from all the various answers, comments, and side room discussions, synthesize them, and post the final conclusion as an answer and accept it, for the benefit of the other users. However, I am not sure if that is seen as appropriate:
(1) Even though the site scripts actively encourage self-answering, my impression from my little experience is that self-answers tend to be downvoted°. If that is indeed so, self-answering defeats the purpose of distilling and sharing. I could pin it with an acceptance checkmark anyways, but it feels counter to the spirit of the approval process.
(2) More importantly, it feels like poaching of off others work and undermining their opportunity for being credited for it. I could maybe address this by including refrences and citing the original answers to give credit to the original responses, but it still detracts from their opportunity to receive approval and votes.
Is there a way to post synthesized answers that is generally accepted? Or should I just not do so?
If the two or three people interested in a question and posting answers work together through discussions about the questionm, exploring the subject matter in depth, they could craft a great, shared answer they all agree on. If the objective of the site is to generate the best, most useful answer, that would be a win. However, there is no mechanism for sharing answers, so in the end, the only way to do this is if one of them posts an answer.
° It is of course possible that it just so happend that the self-answers I gave were more opinonated or of lower quality by chance. And there may be bias due to my small sample size of ~50 answers and only 4 self-answers. But the four self-answers are all near the bottom of approval with an average 0 votes, while the median is more around 3 votes, and the mean around 6; I'm not a statistician but it feels like this is beyond what one would expect under the hypothesis that the distribution is random