Two problems here. First, it isn't the only variable. The amount of variance in the player would probably be outside of the tolerances of a good scientific experiment. Secondly, 'evidences before their eyes' - remember the blue/gold - black/white dress a couple of years back? We don't all see things the same, perhaps we don't hear things the same, and get some bias in our hearing from what we see?
To do this properly, the room would need to be atmospherically strable, a mechanical device used for plucking of the strings, and, most importantly, should be done blind. When someone does this I'll start paying attention.