I suggest one can record the sound in-room with a stereo digital recorder, and compare the sound with a different arrangement or tweak under test. Then compare the two recordings of what the differences are if any, with levels matched between the two cases, of course. That way personal bias is removed, especially if you randomize the recording playback, or name the recorded files with unrelated names to further avoid bias. Then one can say if the differences are real or imagined.You clearly have very high faith in audio memory. I don't. I've been fooled/fooled myself too many times to trust my ability to remember sounds. Taking notes might help, but only slightly - language is too imprecise, and fitting sound to available words dresses the actual sensation in prexisting categories.
Note, this 'test' shows how strong preconceived-notion bias is.
https://ohmspeaker.com/news/sighted-vs-blind-evaluations-of-audio-gear-part-2/

