Edit: this is not meant to be dismissive, it's just that the main problem with social media right now is centralization. I'm simply not interested in jumping from one centralized option to another.
I feel like in a lot of ways, something like Mastodon isn't much better than Twitter, should it somehow take over. Sure you have a lot of servers, but they all run the same software and approach the problem in the same way. It's a decentralized monoculture replacing a centralized one.
How many servers they are run on really doesn't matter. What we need is multiple different paradigms. This, even if it's just one guy running one server, is a step in the right direction.
What I'd most of all like is something that lets the visitors own their own posts while still allowing some form of communication. Dunno, like something combining RSS and a forum.
The 'servers' are simply endpoints within a federated network. You can have a frontend "app" interoperate with those endpoints while exposing a different UX, perhaps one that's more old-school BBS- or forum-like.
Or you can have endpoints that expose a slightly diverging logic, perhaps building on other web standards (such as WebMention) for their analogs of the BBS-like "reply", "topic", "thread" etc. But since these are all part of one Fediverse, interoperability across endpoints will be preserved.
> What I'd most of all like is something that lets the visitors own their own posts while still allowing some form of communication.
This can happen if the visitor owns a server and makes their posts available from there via Fediverse standards. The posts can then either be federated directly (if allowed), or "mentioned" (like a trackback) on other servers.
ActivityStreams is the "RSS" like standard in the Fediverse, which can be implemented entirely statically, with no "active" server-side logic whatsoever. ActivityPub is the active "push" part of the standard, which enables "post" workflows from users; the server itself can also use ActivityPub to "push" its content to a federated endpoint in the network, which removes any propagation delays introduced by polling.