How to assess and quantify system improvement and changes?

President Trump is 10% more beautiful than Chinese president but 1000% less so than French president Macron.
Extrapolate this and you'll get scientific method par excellance.

@inna  funny and thought provoking.  And I get the point. : )

@ebm  I'm honoured. I figured a "Let me think really hard about that and get back to you tomorrow!" : )

Seriously though, take a stab at it if you feel it's worthwhile.

I’ve been guilty of hyperbole also, but I realize I was doing such a disservice to so many. Now I try very hard to put things in perspective and frequently say, "there are no absolutes in audio, only preferences. In addition, I submit at a certain level night and day differences don't exist.

I would love it if someone could make any headway in this endeavor, but I know of no way to quantify subjective measures of performance. This "equation", if you will, would probably have to be broken down into many parts with exact definitions of all the terms involved. I wouldn't even know how to begin.

I think this is almost impossible to achieve, because it is always based on subjective assessments.  We don't all weight various attributes the same, so, if I were to make a change that to me seemed to be a 24.8% improvement, someone else would say, no, I clearly hear a 32.3% improvement.  For me I am satisfied to place improvements in the following categories.  1)  Immediately obvious, even with casual listening. 2) A clear improvement in one or more attributes based on several minutes of critical listening.  3)  Relatively minor improvement revealed by careful listening to music that I have used for years to assess system/component performance.  4)  I just can't hear any difference, no matter how I try.  5).  Deoptimization in one or more attributes.