You are absolutely correct.That would be an invalid test, too many variables.
A proper test would be a standard triangular test with three identical basses except two finished in one type of finish and one finished in a different finish. All three played back to back to a blindfolded listener who must then figure out which one sounds different. If that listener can consistently (in excess of the odds of guessing) spot the different sounding bass then there is a noticeable difference. Any test structured to spot one specific type of finish AND be able to correctly name that finish is basically setting the test up for failure by adding compound variables in the equation. Trust me I used to do this stuff for a living.
Of course that still doesn’t address the real need for a true double blind test where both the player and the listener have no idea which bass is being played. Thus avoiding the potential for the player to play slightly differently on one bass vs another. About the only way to do a true double blind triangular test would be to record a robotic arm plucking a bass held in a jig, assign a reference number to that recording and then repeat with the other two basses. Once done, the recordings would be randomly ordered and played back for a listener who would then need to spot the different sounding bass at better than the average guessing percentage.
It could all be done, but who the heck wants to do it when you know 50% of the people reading the results won’t believe it (stuck on faith based or biased opinions) and the other 50% probably have hearing issues that keep them from ever spotting a difference in the first place.
I say, pick a finish you like and rock the snot out of it, let the audience worry about how you sound in the mix.
I’d be surprised if among those that believe there is an effect, we see a consensus on how a thick finish affects the tone.