Not really. There's reasonable empirical evidence[1] to the contrary: removing large sources/sinks for extremism makes it harder for extremists to unify their message for effective dissemination.
AFAICT, there’s no evidence that online moderation hides extremism rather than eliminating it.
I’m not old enough to speak seriously about the way the Internet used to be, but I am old enough to remember what happened when I would be banned from a service: I would move on, until I either eventually learned my lesson or I was completely banned. Both are acceptable outcomes and the (unfortunate) consolidation of online communities over the last decade has made the former an easier policy target than the latter.
However, those who are truly motivated continue to be exiled into smaller and smaller slices of society, until what you have left is environments of concentrated toxicity, but now hidden from the watchful eyes of the general public.