What is really crazy here is that you are almost talking real sense, sure I get that you are not a scientist as you put it, so avoid attempting to justify your subjective views with fake science, it looks silly.
The issue with measurements is an interesting one, there may be effects in the replay chain that we don't know about and don't really measure but there are things that we can do.
Null tests of various kinds can be performed so see whether a component or process has an effect on the music signal, if there is a change, then further investigations can be carried out and the 'effect' quantified and measured. In most cases the effects are tiny, way, way below audibility.
But, as I am sure you are aware, auditioning components and systems subjectively shows up differences that are often quite consistent and repeatable, why is that?
There are all kinds of reasons why this is the case, level is a big one, deliberate 'voicing' of components another and then there is the way components interact in a system and spending time listening the only really effective way of sorting out what sounds best to you. For most people this is enough, if it sounds good it is good and that is it, but an enthusiast wants more, why does it sound good? What can I do to make it sound better? Etc, etc.
Understanding a little of how hi-fi 'works' and how we interpret it helps the enthusiast sort out those things that are really making a difference and yes, that includes expectation bias and the rest. Simply listening and 'following your instincts' isn't really enough in some cases, a little understanding helps immensely.
Sorry to go on at such length, but if we can avoid the usual cable or blind testing nonsense, this is an interesting subject.