What are we objectivists missing?

On two different occasions, I've read articles by a very famous audiophile who claims LPs sound better than cds, who said something along the lines of, "I really liked how this xxx sounded but I haven't seen the measurements yet so I could have egg on my face." If you like how it sounds, measurements don't matter, but don't try and convince me that it's better than something else.

I figure subjectivists are like religious people and objectivists are the scientists. There is no proof that a new cable, silver fuse, etc improves the sound, but they believe it does As @djones51 said, until you take vision out of any audio testing, there's going to be bias.

"The most reliable method is to listen."

@ghdprentice ​​​@jtucker 

The most reliable method is to listen.    BLIND.  Repeated ABX testing is required.  A consistent score of 95% or better indicates a valid distinction.  All else is entirely invalid.  PERIOD.

Blind listening proves if there is a change or not.  Measurements are unnecessary - we are listening, not measuring.  That removes any suggestion measurements do not measure everything.

@henry53 

"The product measured 0.003% distortion. You cannot hear 0.003% distortion. The same pages waxes lyrical about a turntable that has 0.06% wow and flutter, that's 200 times higher?"

Back to math juniors Henry.  That's 20 times higher.  I note you put a question mark, so perhaps you weren't sure.

I believe there are several missing parts here. One would be time. A measurement is a snapshot. You run a test tone through a circuit and measure what comes out the other end. To measure something like ‘space between instruments in a sound stage’, the relationships between snapshots in time is required.

The room boundaries - discussed often here - are important but not factored in when tying a meter to the end of the wire. Since no two rooms are identical, a change which may help in one room would be the detriment to another.

Another part - the human ear. This is the last stage in sound - a device at the end of every signal path that is never measured. If the mechanism used to hear a sound is not part of the measurement, what use is the measurement? Our ears may react to sound waves in such a way, with more or lease sensitivity, more or less elasticity, that some changes in electrical characteristics of sound are either attenuated or lost altogether. This is like a suspension system in a speaker - dulling out some changes and accentuating others.  The human brain is a really good shelf. It is able to subconsciously shelve noises and frequencies  in our environment to still hear changes in tone while filtering out fan noise or noise from lighting ballasts.  
 

When we test fire alarm systems we do an audibility test. The recording which is played through the speakers is loaded into a hand held device. You walk throughout a space with the hand held and it compares what is being heard to what should be being played. The purpose is to determine if the words can be made out, but I think you get the idea. I consider that type of test to be closer to a real world test and you’ll find issues that cannot be identified by measuring the signal, ambient noise, or the speaker. I think the impossibility of doing the same in our listening rooms, with wires to the back sides of our eardrums, is why we reserve our judgement for a final listening test and put the results over any other test. 

You mention cables,  if 2 measure "exactly " the same then they will be indistinguishable in an ABX test. The reason people claim they hear differences in components that measure the same is due to confirmation bias. 

Another possibility:

The new cable is different in length, gauge, or termination than the old cable.  So it brings slightly higher gain to the system, which some users perceive as improved SQ.