I am very aware of the kind of centralization you describe. It is inevitable. And will happen to Lemmy, and Mastodon, and Matrix if/when they become large enough. It’d go so far as to say we are already there with the World Wide Web. What you describe is simply the Pareto Principle in action and it’s a law of natural order.

For me the measure of success of a decentralised service is a bit more modest: I need not that the market shares be evenly distributed among parties (that is a discussion of economics I frankly am not all that interested in debating).

But everything else still applies today: I can spin up a mail server today for pennies and start handing out my email to Average Joe, and (most of the time) they won’t roll their eyes and say “uh how do I use this?”. This is the ultimate, practical test of success of a decentralised communication service. For me it need not be fancy, it need simply to be used.

I am aware that in some developing countries this is not entirely true. I also think that these communities haven’t had the time it takes to learn the very expensive lessons of centralization, as I made in my original point.

I also think some backsliding is possible, and probably happening with the marketshare of email providers today. Then there will be come some outtage, companies will lose profits and we’ll all collectively relearn that maybe putting all our eggs in one basket isn’t such a great thing.

If there is two things you can count on in people, it’s greed and forgetting history.