Your threshold number is mostly an artefact of the test that made it

Change the stage length or the starting workload and the same athlete on the same morning will produce a meaningfully different result.

Two labs, one athlete, one morning, two different threshold figures. This is not a scandal. It is what happens when you measure a moving process with a staircase and then report which step you were standing on.

The incremental test asks an athlete to work at a fixed intensity for a stage, takes a blood sample, then raises the intensity. The reported threshold is the workload at which blood lactate reaches an agreed concentration, or departs from baseline, or turns up on a curve, depending which of a dozen definitions the lab has adopted. All of that is sensitive to how long each stage runs. Short stages do not let lactate equilibrate, so the reading lags the effort and the threshold comes out high. Long stages allow accumulation and drift, so the same athlete looks worse. Three minute and eight minute protocols can differ by a lot of watts.

Then there is diet. Turn up glycogen depleted and the lactate curve shifts right, which looks like a fitness improvement and is nothing of the kind. Turn up after a heavy carbohydrate breakfast and you can manufacture a decline. Athletes learn this. Some of them learn it deliberately.

None of that makes the test worthless. It makes the number a local measurement rather than a physical constant, and the correct use is longitudinal, in the same lab, same protocol, same time of day, same nutritional state. Compared that way it tracks real change beautifully. Compared across a training group, or quoted at a coaching conference, it is close to meaningless.

My complaint is about how these figures travel. A threshold power gets written on a plan, the plan gets shared, the athlete changes coaches, and eighteen months later somebody is prescribing sessions off a number generated by a protocol nobody remembers. It has acquired the authority of a measurement while losing everything that made it interpretable.

The field alternative is better than its reputation. A hard, honest, thirty minute effort on a known course, repeated every eight weeks, gives you a functional intensity that already contains pacing, heat, terrain and the athlete's willingness to hurt. It is noisy in the way athletes are actually noisy.

I would rather have a rough number from the world the athlete competes in than a precise one from a room with a fan in it.