I'm sure the one who did the test will give details as he has here before,
if measurement includes a listening alone test of a serially tweeked
system of good repute, as to it's gear, and one made as unlike it as is
practicable, then this notion can not be confirmed. The highly tweeked
system could not be distinguished from that set up to violate as many
tweek guidelines as possible.

If you have applied several different tweaks, the final output will be
exactly the product of each individual one and will be measurable after each
changed component.
Your argumentation is not valid, it is governed by your belief and utterly
unscientific.

--
ciao Ban
Bordighera, Italy