I don’t want to disappoint you with reality but nostr hasn’t won yet. And there are a lot of protocols created during last decades, many even more censorship-resistant, did they also win? If I create the same protocol as nostr but with all keywords capitalized or in spanish does it also win by default?
The problem with what you call censorship is complex because everyone views it in their personal way :) which is kinda incompatible with “global discourse”
And you also want moderation. Now you call it “mute and paid relays” and looking forward for “client algos” then you will call it “topical relays” or similar
There also will be people who will kinda get as close to “no moderation” as possible - by using only private/family relays (which is a moderation by social graph property under the hood)
So, moderation will always take place in something bigger than monologue. The question is how to take control over this moderation so it would match your needs as much as YOU want as possible instead of the platforms CEO AND keep room for global-or-similar discourse