Interesting comment, John, about repeatability. That is one of the ways scientists check each other's work, as I am sure you know, but I mention for the sake of those who are not familiar with scientific process. Thank you for bringing it up.
Ervin, on the other hand, did not write the books so that readers could do knock-offs of his work. He seems rather opposed to makers copying each other and neither his books nor his videos offer whatever it would take to make a knock-off. There is good precedent for this, just as there is precedent on the other side of the teeter-totter too (Martin revels in those who would copy them). In any case, I am not surprised that you could not repeat his process, yet found the books valuable.
Myself, I could say about the same thing. As the books teach me what to look at and what to listen for, I don't always reach the same conclusions the author does. A specific case in point is coupling tops and backs to ribs. I like both plates COMPLETELY decoupled, whereas Somogyi instruments generally display at least some coupling. So it is a case of my using his framework - at least my interpretation of it - and reaching a rather firmly felt, but quite different end point. I think that is well within the intention of his writings. As I say, there is a lot of Socrates in Ervin Somogyi.
But to the issue of repeatability, but from a different direction. Ervin himself is quite good at repeating his process, judging from the recorded sounds of his instruments that I have heard. Just as the pre-war Martins have a boomy, cutting sound that projects a specific aspect of steel string energy, so do the Somogyis push out a typical resonate, thoroughly discharged bass that I hear from one instrument to the next. (Wish I had heard some in the flesh, but have not. Such is modern life and high powered sound equipment.) So if Ervin can repeat it, that is what ultimately validates it. Secondarily, if those who find it useful in some way or another, continue to find it useful, that adds to its value.
Science can generate very good instruments, but not great ones. It takes a specific human, some luck, and a lot of work to do that. This is a belief, guess, intuition - not anything I can or would attempt to prove.