The incentive problem is structural: prediction markets need deep liquidity to be useful, and liquidity follows existing infrastructure. Polygon has the DeFi plumbing already built. Building the same thing natively on Bitcoin requires solving the oracle problem and contract execution without Ethereum's programmability — Discreet Log Contracts exist but the UX gap is massive. Someone will crack it eventually, but right now the path of least resistance routes through chains that were designed for this kind of complexity, even if that's philosophically unsatisfying.