Uh... I'm sorry to say that you do not appear to understand how power meters and the corresponding mathematical models work.

Power meters measure the exact power applied to the bicycle, i.e. they're measuring the actual torque generated by the rider, not "how many calories the rider is burning" or "how many watts the rider's body needs to generate." They are usually installed in the crank, bottom bracket or as the rear hub. So, when we talk about a rider applying an average of 200 watts over 20 minutes, that means that the pedals are receiving 200 watts of power from the rider.

That's why a power meter is a more objective measure. It doesn't matter who the rider is, how good or bad their aerobic system happens to be working that day, or the temperature, or whether you're climbing or descending, or whether the rider has more or less mechanical losses for some reason, or if you're sore (or if your blood glucose is low) from the previous day's efforts, or if that yogurt you had that morning was too close to the expiration date, and so forth. The power meter doesn't care; it's just measuring power that is actually applied to the bike. Thus if the rider is applying 200 watts to the pedals, they're applying 200 watts to the pedals and that's exactly what the power meter records.

The Kreuzotter and other mathematical models are also using this as their measure. Thousands of riders are comparing their power meter results to these calculators as well. As I may have said before, obviously it will be simplified compared to the real world, but it's generally good enough for these types of questions and would be corrected if they were off by orders of magnitude.

By the way, the actual measure is of "watt-hours." Going from an average of 200 watts over 2 hours to 201 watts over 2 hours really isn't terribly significant -- unless you're in a race or, per staephj1 and invisiblehand, an extremely competitive club ride.