What self-proclaimed objectivists are missing is:
The majority of perceived differences between cables and components are SUBTRACTIVE. And steady-state measurements can only measure ADDITIVE "distortions."
Not surprisingly components that measure best with steady-state signals are most likely the ones that subtract the most information.
If you measure using real music signals (into real loads under dynamic conditions) it is easy to see what has has been removed by the very design strategies that give good results using sine waves.
There is also the issue of loudness compression; which as far as I know is the number one reason audio does not sound like what was happening in front of the microphones and it is not measured at all.
herb