At some point though we have to acknowledge that nostr doesn't scale for the intended use case. It's cool, but it doesn't scale. So the only fair comparison is between Nostr and other architectures that are also innovative but also can't theoretically scale. If SSB were still around it'd be a fair comparison. I'm sure there are some others. Pubky, Keet and such are a class that *can* theoretically scale. The chance that they will isn't all that high in light of the history of the internet in general, but there are no fundamental technical barriers preventing scale (whereas with Nostr there are). That said not being able to scale is not a bad thing. Nostr architecture could form the basis of many little worlds that don't connect to form a big world but still makes Nostr a useful contribution to the web.

Slight tangent here, David, but I think this approach works in the religious sphere as well. Orthodox enough (within consensus rules) not to be excommunicated but free spirited enough to question what needs to be questioned. “nostr … leaves questions of consensus up to custom implementations, in both client and relay”