Spam isn't (generally) illegal, and yet virtually everyone would argue all platforms should moderate spam because to not do so massively degrades the service quality for everyone.

I mention this because I think content moderation, as a debate, should not be framed as whether or not to do it at all -- but rather as a recognition that deciding to moderate a particular class of content carries a cost-benefit tradeoff.

It may ultimately be the case that we decide that a certain class of content is not worth moderating, because when doing so the costs outstrip the benefits, but let's keep that as a pragmatic discussion rather than a principled one about "all" speech.

Your instinct might be to respond with "of course, spam should be moderated, but besides that..." but I think the knowledge that there could be some series of enumerated exceptions suggests we should stop pretending they're exceptions at all.