The problem is developers pushing a technical choice of protocol with little regard for user experience. How does the XMPP protocol improve on the user experience? It doesn't unless you have a client/server implementation that can be deployed and showcases that power. Think SMTP and email. You want XMPP to work, find a simple way to showcase they power.

I've run the original jabberd server for an organizations internal chat server along with all the open source bells and whistles that go with it. Let me tell you I didn't once see the value of that except to save money and for privacy. But it was an organisation. If we want open federated communication the standout thing has to be the ease of use, the quality of experience and to actually highlight it's benefits. Right now I don't see that.

What you are describing is basically the gap that Snikket aims to fill - although for personal communication, it's not particularly aimed at organizations (I think the requirements for that are different, and there are many existing options already).

I know at least one project aiming to fill the same gap for organizations though (again, using XMPP), but pre-release currently: https://prose.org/

I'm a firm believer in essentially what you said: The protocol doesn't matter, it's all about the products and what they can do and how easy they are to deploy and use.

I'm sorry but if your point of data is the original jabberd server then you are not up to date with the experience you can have today. And the experience today is good, whether you use Prosody, Metronome, ejabberd and probably others, and when you chat with Conversations. You have all the bells and whistles of social networking with Movim.

It' s not necessarily a better user experience but it provides a far better user independence, and that definitely counts.