Most of the time if your question is mostly a duplicate but the other question (or its answer) is missing some critical factor, then you can appeal and say, "I don't think this is a duplicate of <x> because <y>" (or modify your question with the stipulation that the answer solve the critical factor) and the duplicate marking will be removed.
The other thing is, even when questions get marked as duplicates or closed, half the time I still see them accumulating their own answers, so really, what's the big issue?
As an aside, my only experience with having a question marked as a duplicate happened on a sister SE site - it was my first post there, I asked a question, then after posting I found the answer in one of the "related questions" links on the sidebar. I answered my own question with a link to the other question and a description of the answer I found there, then flagged my own post as a duplicate. Both the question and answer were highly upvoted.
I'd be interested to see some statistics. I strongly suspect that when most newbies are told to fuck off that they do, in fact, fuck off.
As for statistics, you may be able to find some by querying the data explorer: https://data.stackexchange.com/stackoverflow/query/new
I suspect it would be difficult to (automatically) distinguish between instances where a user was discouraged by the message and instances where the user's question was answered at the link. Perhaps the fraction of instances where the question was edited after being marked as a duplicate, then re-opened, is an OK metric for how willing a given community is to re-open questions.
Might be interesting to see more recent statistics, but it still gives an idea.