For example, let's see this question from ELL: "I have no question"--> "Me, too" or "Me, either"?
Let's call this A. Question A is on top of my Google search, currently has 4 upvotes and considerably well-researched. However, it is marked as duplicate to question B, which is closed for being poor-researched. A's answers attract more votes too. In short, we want A, not B, regardless of their history.
Should we reverse this situation by saying B is a duplicate of A? That way, A is now opened and new answers can be written, while B is still closed. When I head to A, I see that it's duplicating B, so I click B with the hope that B is better than A. It's not. After I find an answer elsewhere, I want to answer them but I can't anymore. Without looking at the revisions, my thought is that B should be a duplicate of A, not the opposite. I think if duplicating questions has more interacts than the duplicated ones, especially with new edits, then it should be added to the review queue.
So, if a duplicating question has good quality, and the duplicated question is bad, should the relation be reversed?