I don't know the answer in this specific case. I expect the answer is most likely no.
From a scientific standpoint your concept of what would be an ideal situation is quite sound.
From the practical standpoint of working with human beings and trying to develop the best methods, generally speaking trying to do that sort of thing slows progress so much that it's not worth doing it.
For example, Coach Thibaudeau's time is limited. It would be a desirable thing to have ALSO had him training a large number of subjects -- sufficiently large to meet statistical concerns -- that didn't use the protocol, for comparison purposes. But does he have the added time? Or does the entire project get run into the ground in the process.
As he and others have been experimenting with the nutrition protocol for quite some time, it was already known that it makes a major positive difference.
So the goal instead is to put together what is understood (by those doing it) to be the best they know how to do and see how that works.
Then maybe at a later date see if it can be quantified how much each element contributed. But not hold everything up on determination to establish that before having anything.
Now, I don't know that that is the case here. But it is what I would expect, having accomplished things by both methods: science done to published, peer-reviewed journal specification, and a scientific approach that makes practical advances but doesn't get bogged down in things not necessary to determine if what is thought to be the best approach we know how to do does work and if so, how well does it work.