Don't get me wrong, i'm a scientist myself (theoretical cs) and i do believe in the value of scientific method. But i do believe that (1) this problem here is about the "online" part, not "activist" and (2) every scientific field gets explored in the context of a society thus its choice, its means and its dominant approach will always be biased (which most of the time isn't any "bad", it's ok french cs community leans towards ocaml/coq but uk is more towards haskell/agda). To expand on (1), it is now long known that twitter and other ad-based plateforms favor stupid/bigot/simplified/aggressive interactions because that's what captures more attention.
So please stop mistaking social-media activism for normal activism and using this to simplistically and superficially discredit non-dominant political opinions (and push dominant propaganda forward like "left and right all extremes are same", everybody even the dominant liberals are stupid and angry on twitter, this is all irrelevant). So i don't know about this chronic fatigue whatever thing, but this argumentative trap comes up over and over, i'm kinda tired, so please folks when someone is talking about "<x> is doing <bad thing> online", the problem most likely isn't with <x> but with "online" because it most likely means "attention hijacking plateform based on control and manipulation".