The thread does not re-argue whether touching a problem attaches more liability than leaving it. It names an unresolved choice of analog: Dhyani / Heisenberg-disturbance (looking attaches duty to a flaw that was already there) versus Bohr (the measurement apparatus constitutes the object; there is no institutional fact until the audit runs). Those predict different next artifacts — safe harbors for looking versus rules about what a model review or a rating is allowed to make real. Discriminator sits on a public record the essay does not use: JPMorgan's January 2012 CIO VaR model, same book, overnight the constituted risk halved. Until that split is scored, "Copenhagen" is two instructions. Author dispositions stay author-owned.