This is a shortsighted take IMO. "Fringe" websites are growing rapidly while engagement on the big, normal websites is declining or slowing down. The problem is people want to talk about these things, and if that's what people want to talk about they're going to go where they can talk about them. This is directly addressed in the article (when we say we want to ban hate speech, what we really want to ban is hate, but hate cannot be banned). These big legitimate sites only feel permanently dominating now, but are not permanent by any means. People don't go on the internet and talk because talking is a way to use Facebook, people go on Facebook to talk about what they want to talk about. If people want to talk about edgy or even hateful stuff, they will, and if they can't do that on Facebook they'll do it elsewhere.
That's not to even get into the fact that all to often these terms like "fringe websites" are used by entities to shame users into not using their competition, which further degrades the integrity of any anti hate message.
I don't know about engagement on mainstream sites declining, but the rising popularity of sites like Parler and Gab seems undeniable. Parler's recent (attempted) deplatforming is evidence enough: you don't bother with a tiny fringe site that isn't growing or wielding any influence.
Better question:
If increasing, is it because more topics of conversation are getting binned as 'fringe'? Or are extremists merely circumventing centralized efforts at blocking by moving around, which is exactly what those anti-censorship weenies like me jave been saying will be the inevitable result?
The problem you ignore or quash never goes away. It just gets harder and more distributed.
That's not to even get into the fact that all to often these terms like "fringe websites" are used by entities to shame users into not using their competition, which further degrades the integrity of any anti hate message.