Forget the BS instances argument everyone keeps having.

The question I ask the ATProto folks is this: Explain how, if Bluesky went out of business tomorrow, the network keeps running without disruption.

I know that it can keep working, but unlike the Fediverse, it won't happen smoothly like if a major instance shuddered and people had to migrate. Bluesky holds the whole thing together in a way that impedes good decentralization.