The fundemental problem here is that SO was designed to answer people's questions, and now has so many questions and answers that it is a kind of wiki for coding question.

Answering questions doesn't just help people, it is also very important for SEO reasons. The more content that gets added, the more SO's pagerank algorithm is increased for all the terms mentioned (though there is probably an upper limit somewhere, there's no reason to assume it is reached). There's also no telling what keywords the post contained that might find the top spot on a Google search. And not just Google search, but SO search as well. Although a question may have already been answered somewhere, it nevertheless might be missing an important keyword that the other post did not have. I'm not saying that that is the general case, but there's no way to know it.

Many "duplicate" posts have the top spot on a Google search. If they were removed, would another post take its place? In most cases, but certainly not in all.

So don't imagine that SO is ever going to drop its site ranking by removing these great sources of trafffic. Its just not a fair thing to ask. I'm not working for them, just pointing the obvious. They appear to be sane in that regard.

Another thing that seems clear is that in the tags that have been around for a long time, there cannot be enough real questions of the "nobody has ever asked this or something like it before" quality to satisfy the number of people who want to answer questions and spend time on SO. So basically, what you have now is people who are shifting roles from expert question answerers to "wiki site maintainers" who are spending time closing questions and asking users if they have done this or that.

So, in short, I would say that although there is merit in cleaning up the site, for SEO reasons it is totally a wrong premise to think that SO can ever be a wiki site. This is not just driving traffic, but how search works. The question diversity is too large to meaningfully categorize all of them together, let along the fact that no search engine exists that can search categorized data as well as google churns through the "web" via keyword indexing.

The Noob problem could be fixed easily by putting up real barriers to entry, but no one is going to do that because... well... it seems that the activity has merit in the sense of driving site usage.

So - one possible solution - put up an actual wiki site and link to it from these duplicate questions. Maintain the wiki site, and when a question comes in, offer the wiki link as an acceptible answer (ie. an answer that the OP can accept). Add preview for SEO. Then add an official notice at the top (by moderator if needed) suggesting the approved wiki answer "Find up to date and maintained information about this question topic here". Now the question has been classified and it has real SEO connections.

As to user incentivization, isn't the real question "what do you want them do do"? You can't detect bad questions without a human moderator, so that is just a waste of time. You can't exclude beginners from the site without testing them somehow or making them pay money (neither of which are reasonable).

So - if you want to make a wiki site: make a wiki site.