My previous question What to do with useful answers to questions closed as duplicates (or by some other reasons)? got an answer which satisfied me entirely: if a duplicate question has got a nice answer that should not be lost by any means, moderators may merge the questions.
However now I seem to have exactly this situation - my recent question Are there nice isomorphisms $\operatorname{S}^2(k^n)\cong\Lambda^2(k^{n+1})$? is an obvious duplicate of Coordinate free isomorphism between $d+1$-dimenional antisymmetric rank $2$ tensors and $d$-dimensional symmetric rank $2$ tensors, as noted by Robert Bryant (well, more precisely he did not claim it is a duplicate but I believe it most clearly is).
In fact, even the generalization that I mentioned in the end seems to already have been addressed, as explained in a comment to an answer to that earlier question, https://mathoverflow.net/a/42859/41291.
Before I knew all that, my question got a truly beautiful answer which, although related to answers to that other earlier question, still, I believe, complements and enriches them essentially.
So it seems a situation when a merger should occur. How do I proceed? What are the criteria? Or this is not for me to decide?
Let me add that I hate to create extra work for moderators, and I promise not to misuse the fact of a precedent, if any. By the way, have any mergers occurred before? Is there a way to detect them?