DBT is subjectively determined. How would that fit in a mathematical formula? If you are going to accept math and science, DBT has no purpose. If you are going to accept subjective opinion from DBT, you have to accept subjective evidence countering it. If not, I call BS.
The great value of DBT for audio is that it can clearly demonstrate (in many cases) that people who say they hear a difference, don't.
(As in the well-known DBT where self-described 'golden ears' listened to a $10k+ Futterman system and a $200 Pioneer, and stated that they were certain they could distinguish between them in a DBT given immediately after the listening session. Pre-DBT, the 'golden ears' also said that the DBT was well-designed and should give valid results. The result: 'Golden ears' often couldn't tell the difference between the amps, and where they did hear a difference, they mis-identified the amp about 50% of the time. i.e. results no different from chance. When later informed of their own results, the 'golden ears' frequently changed their opinion of the validity of the DBT, stating that it was not a valid method. )
DBT is one of the standards in clinical trials, so I don't understand how there is a conflict between DBT and science/math.
A couple of years ago, when I got interested in the 'audio hobby' and building amps/speakers I built a 'cap switcher' with a wired remote control. Tested a number of 'high-end' caps against Russian PIOs and off-the-shelf cheap poly caps. All were good, functional caps. Caps were inside the case, and the selector switch was unmarked- ie the listener didn't know which cap was selected. Colored LEDs on the faceplate allowed people to state which combo they preferred. I also put duplicate sets of some caps into the order. 8 selections offered. None of the guests (a couple with deep audio interest) to my home could tell much difference between the caps, and none of them could identify the duplicates. Results strictly random. Follow-up: I offered to lend the cap switcher to local audiopliles for a get-together. Not surprisingly, nobody was interested.
You should build something similar for yourself- it's interesting.