You're right, i've missed the point

I thought it was about the music, not the hardware.

hardware >>>>>>> music

'audible differences'>>>>>>>'better'/worse music

I don't see the point, unless one just wants to tabulate abstract figures in peculiar listening scenarios for the sheer thrill of it. In which case, why not buy a chemistry set and a few tools, rather than hifi equipment.

And how exactly do you go about regulating The biggest variable in a double blind listening test?

 

What exactly does 'reliable' mean?

Absolutely nothing.