Just a few db differences can allow for separation in timing and level for the given individual ear. Each ear being different.
So diffusors can work for one speaker, and room, and not the next. then we introduce different ears and we've got 3 variables that dictate the seeming quality of a diffusor. As all three have to be aligned for it to all calculate out, but none of the three variables can be accounted for.
Then the electronics, wiring ac quality, and so on. Actually, 4 uncontrollable permanently variable, er, variables.
My predictions, if I have any... on the efficacy of diffusors... is in no way a slam dunk for the next person.
A general pattern emerges, yes, under the umbrella of the four variables. But no hard number will ever emerge.
To argue this, well one better come armed. :P