Wow, somebody is misunderstanding the whole point of this exercise.
Swamp is trying to explain the difference in results between what we would expect from the stated specs and what we see in the "real world" tests. If you can't predict performance from the specs then the specs are wrong (EDIT: or the model is), one way or another. Science is based on predicting what will happen and testing those predictions, you know. But please, post more of the "it's close enough, why do you doubt it" stuff.
You can predict performance in the real world, you know. It's the basis for car design. Models. So far the CarTest model has worked for other cars, why not the IS-F?