Earlier quoted context omitted.
I was around at the time, too, and you're missing that the early Internet wasn't free at all: the great majority of the users were part of some postsecondary educational organization. Those places strongly select for very specific kinds of people (smart, collaborative, curious, reasonable, not murder-y), and imposed both formal and informal controls on what was said. Your notion is that 3 corporate-run, lowest-common…
Maybe we remember different things. The way I remember it, the early internet had plenty of horrible stuff and the sky didn't fall. I think silencing people (legally or otherwise) doesn't help reduce murder rate or anything. It just silences people.
You might think that silencing people doesn't reduce the murder rate, but you would be wrong. There are specific crimes of speech related to that, like incitement to riot, because that kind of speech historically leads to violence. You can also look at the history of people who end up killing people. Many of them have histories of radicalization, and that path starts with the mildest of entrypoints. See, e.g., Bellingcat's examination of the radicalization of internet fascists: https://www.bellingcat.com/news/americas/2018/10/11/memes-in...
Or look at how Bowers and Sayoc got to the point of violence: https://www.bellingcat.com/news/americas/2018/10/31/magabomb...
I agree that silencing people in general is bad. But it's very clear historically that some kinds of speech if tolerated lead to violence and death. If you think you can demonstrate otherwise, by all means take a swing at it.