Testing: does individual skills matter?

[Toxcatl]

A study case i would like to see

- I bought a nordgen f.t./ academy / recommended (all those players are nordeus generated with untouchables skills and almost equal the three categories).
I made some training to him and he has high 6* quality (115% for example).
He is playing as ST but doesn't score much, 1 goal every 2 or more games (we must not choose him as a F-K or penalty specialist).

- The second season, I train him mostly in scoring-finishing skills which increased a lot.
As a result, I can see that he 's scoring a lot now.

To make a research we need to get rid of the influence of other factors.

First of all, the team is changing, so the guy won't be playing with the same team all the time. You could not step twice into the same rivers. Maybe he didn't score because passes/crosses from AML/R were bad, but on the next season this situation will change.
How exactly do you want to train him on the first season? Without training Shooting and Finishing? What about other Attack skills?
His age will change. Many managers say that 18 y/o don't play with their full potential, maybe it's just a legend (I haven't noticed such a thing). But age may have an influence on player's performance.
Two players would solve this problem, but, obviously, their internal programming would be different. No way either.

In other words, we can try something like that, but I don't see how we can separate flies from the soup in this case. The results won't be trustworthy. Maybe I have an idea, but I need to think more about it.