To be honest, I agree that Blossom adds complexity without a clear payoff. I’ve talked to a lot of devs in the Nostr space, and most seem to implement it simply because others already did—not because of any real performance or architectural benchmarking.

I’m not saying “use file drop instead.” I’m saying Blossom doesn’t actually do anything special, and there are already dozens—probably hundreds—of better, more mature options available.