When all social communication happens via app on one or two platforms, and a dozen people have full access to every private communication for billions of people and the ability to decide what communications you can and cannot transmit via the primary and sometimes exclusive methods you use to speak to your friends and family, an extremely dangerous situation results.
It is within the power of Facebook, for example to selectively allow or disallow discussion of COVID/nCoV based on user, region, geolocation, location history, et c - even in DMs, even on Instagram or WhatsApp.
Imagine if a regime insisted that they censor a specific term or set of terms within their country, or be blocked GFW-style.
I doubt it’d even require that much new code. They already have DM censorship mechanisms in place to combat spam and other forms of automated messaging.
When you have no other contact info for those with whom you communicate, it gives them a point and click information blackout ability, even for “private”, person-to-person communications.
Everything from protests to genocide can be made impossible to discuss, and thus invisible.
They have full location history for a lot of users, too. Imagine them temporarily shutting off DMs for every user who has ever visited a specific location for the duration of some national security emergency specified by a government. Only temporary, of course - but suddenly the protesters of the concentration camps can’t send or receive any messages from anyone. Nobody else notices or is affected, so it doesn’t even make the news.
It’s an extremely dangerous state of affairs. I don’t even think Facebook realizes what a huge menace they have inadvertently produced, if for no other reason that I don’t think the national military in the country where most of Facebook’s SREs reside has pushed the issue with machine guns yet. Those guys don’t mess around.