Is it a problem if specific users harvest highly-voted questions on SE, then self answer, to farm a huge amount of rep? And if so, what can we do to prevent it?
What you mean is that people basically could loosely copy content from elsewhere giving attribution if necessary. That new content was not already existing here but gets highly upvoted and therefore generates lots of reputation.
I'm not sure if it's a problem but copying (even adapted) content from elsewhere surely is rather a repetition and does not add much to the overall existing amount of knowledge. Therefore it might be a bit of a waste of time.
Then again, if it's highly upvoted, it means that it's very useful, so probably something we would want to encourage. We want lots of highly upvoted ontopic questions.
We probably couldn't prevent ontopic questions derived from content stored elsewhere unless we insisted on absolutely original questions every time, i.e. questions that are new not only to this network but new to the world as a whole.
That touches on what Codidact or the communities that form Codidact actually want to achieve? Be an even larger library of knowledge than already existing elsewhere? I couldn't yet find information about that and will probably ask about it one day.
I haven't seen it happen, but asking this question on a coding site seems like a pretty straightforward way to guarantee yourself a lot of rep with minimal work
Even if this hasn't happened yet, I'd be surprised if it doesn't
happen in the future
Looking at software development CD it hasn't really happened on a large scale. It seems that even though it is not a big amount of work, it's still not very appealing to users.
My take is: Reputation is supposed to be less important here, so we do not really need to worry about that. The real worry would be needless duplication of content, but if people want to spend their time doing that (and give proper attribution if needed), why not. The content itself always was free and not thought to be stored at one place only. But on the other hand, search engines seem to think that one canonical place for each piece of content is kind of enough (they wouldn't have to, they could also just randomly alternate between all existing copies of an information).
Therefore it does not need to be encouraged, but also not really suppressed. Indeed for a selected number of Q&As it may as well make sense (the most often used ones as they will be asked anyway and are ideal duplicate targets). There is a kind of gold rush opportunity out there for the early adopters of Codidact to create lots of highly useful knowledge, but then without reputation, all you would get is the sense of having helped somebody else.
If I have some free time, I'll try to do a bit of that, at least asking some of the more important questions, but I don't need any reputation for it. I'm happy if other people read about it and find it useful.