we all draw the line in different places, but I think most would agree that some subset of user-submitted content is purely abusive. if the only recourse available to the victim requires identifying every individual poster, it's effectively impossible to stop.
imo some sort of safe harbor mechanism is a reasonable compromise. we can't expect the host to prevent every single instance of abusive content, but we should require them to make some kind of reasonable effort.
I'd say Kiwifarms has failed to make that kind of reasonable effort.
There is a good faith moderation clause in part of section 230, but I'm not sure how it has been enforced in the past or if it is similar to what you are thinking of.
You just hand wave all the difficult things away with the weasel words, but the majority of the population, in reality, in no way agree on the actual definition of any of the things you're referring to. Not what constitutes abusive content, nor what should be done with it, or whether anything even should be done with it.
yes, I'm being intentionally vague about "abusive content", because that's not really what I want to debate here. revenge porn might be a good example though. what fraction of people do you think would be okay with having their privately-shared nudes posted over and over with no practical recourse or mechanism for slowing it down? no one I know in real life would accept that.