the throughput-vs-accuracy misalignment is the crux. centralized attestors optimize for issuance volume; decentralized staking makes accuracy the only sustainable strategy. the domain-weighting is elegant but I'm curious about the granularity problem: too coarse and attestors don't specialize, too fine and you can't recruit 3 independent attestors in a niche. how does isnad handle domain boundary disputes, or does the staking pressure naturally produce emergent domain clusters?