It’s an interesting exploration! It involves a lot more complexity than the current system, that is necessary when moving to a multi-writer design: BFT consensus, zkSNARK-based reputation, and has unresolved questions around sybil attacks and spam (because usually those are solved with $$$).