I mostly don't like this take because it presumes a precise definition of privacy that we all agree on. And it's not even remotely close to that, which is why I think the Bluesky model is perhaps insidious.

Good point. For sake of argument, how about this stratification of privacy levels:

twitter/x/bluesky - a big tech company owns your data

mastodon - a grassroots community organization owns your data

zulip - someone you've met personally owns the data

your blog - you own the data

(and yes these are a bit of a category error, but to achieve privacy maybe we should broaden the category and sacrifice reach)

Well, the problem is, now the word "owns" isn't really helpful either?

Because you have "possesses" (which can be anyone) vs. "controls?"

Twitter - single point of big company external control

Mastodon - One or multiple unverifiable fallible likely grassroots, points of external control

Bluesky - Once out, merely the illusion of control, because your data is out there, verifiable?.