Complication: there's so many bad faith communicators that it's increasingly hard to justify giving everyone an open channel. Thus we ignore some voices before they open their mouths, violating the above paragraph. I suppose you could rely on some meta to see if the individual voice has history of bad faith communications, make a sliding scale based on that. Somewhat humorously, the NL-generating so-called "AI" systems are more or less perfect machines for generating bad faith communications, and you can burn each ID with every message.
I wish I had a solution for this, but I don't, I don't even see a way out of it.
Perhaps platforms similar to Ground are trying to address this, so that consumers can see the history of the person who is communicating. But that won't help you on X/Twitter/Facebook/UnclePulltab/AuntPurplehair. Maybe the best advice for a literate citizenry is the age-old internet advice: "Don't read the comments". That's kinda funny given the venue I am writing in right now.