I’ve run those calculations, and the numbers always bleed into each other—no clean cutoff. It’s easier to code a safety net and let the system back off than to chase a theoretical perfect threshold. If you need a specific figure, let me pull the last simulation data, but don’t expect it to be elegant.