I'm very interested in 'new things', especially in this area, and I find it incredibly frustrating how bad they often are for a significant number of users.
One of the more promising initiatives I've seen recently is Patchwork, which is a kind of Facebook or social network that runs on the Secure Scuttlebutt (SSB) protocol.
It's a user-friendly app that you can download, and after getting over the hurdle of adding a 'pub', you can get things going pretty quickly.
What frustrates me is that I can imagine quite a few people would go through the trouble to install the app, non-geeks included, even though it means going to a weird .nz domain and even though simply searching for 'Patchwork' on Google isn't enough. In my experience there's a subset of the general population that will be willing to do this.
But then they get an app that does nothing, and they'll have to figure out that to use the app, they need to add a 'pub' server. Doing so requires a bit of searching and then adding the pubs to the app. I'd say that's a bridge too far for most.
The same goes for developers. Patchwork is written in Node.js and for various reasons there's no easy way to write clients in other languages. Even though I am familiar with Node.js programming, it's enough of a hurdle that even as a fan I just resigned myself to just wait until something more user-friendly shows up (Elixir client plz).
My point is that Patchwork/SSB has been one of the best-implemented things I've seen, and yet it sucks enough that I hesitate to recommend to 'regular' users or developers. And the frustrating thing is that, considering the effort involved, it wouldn't be that much more work to bridge the gap. But doing that is just not interesting enough, and so the whole thing remains a niche that will probably die in due time.