>The question hence becomes 'what point does a site or service go from being innocent to exploitative?'
>Few would say Wikipedia is exploiting its users. Some may say that about StackOverflow or YouTube, but many would say it about BuzzFeed or the Huffington Post.
You may have answered your own question with your examples. Wikipedia isn't monetized at all. It relies entirely on fundraising. All the rest have some sort of monetization model which is built around the assumption that their users will generate free content that gives their platform its value.
On the monetization spectrum, though, I would say StackOverflow does much better than YouTube. It's designed entirely around providing value to its users, rather than merely selling their eyeballs (you can find lots of information about SO's approach to monetization here: https://stackoverflow.blog/2016/11/15/how-we-make-money-at-s...).
My personal spectrum from best to worst looks like this:
1. Funded by donations (sort of utopian and doesn't always work)
2. Funded by directly selling services/goods worthwhile to the community (pragmatic and mutually beneficial)
3. Funded by selling generic ad space to third parties (lazy as business models go, but newspapers have been doing it for centuries - this is usually a sign the organization doesn't have a well-defined audience: cf. Facebook/Youtube vs. Stack Overflow)
4. Funded by secretively profiling people and selling their personal information without telling them (just evil and gross)
As for platform vs. publisher, that's been an interesting topic in internet law. See for instance Cubby vs. CompuServe in 91: https://en.wikipedia.org/wiki/Cubby,_Inc._v._CompuServe_Inc.