power response, directivity... etc....etc... all come into play and any variation no matter how minor will result in differences in playback.. that's why we have a million flavors of audio goodness and is why there never will be the perfect rig nor a group of people that will agree or like 1 said application...
that's the beauty of audio 
Randy, don't you mean the beauty and the beast?

My favorite example is two identical loudspeakers with identical (or as close as can be to that unachieveable goal) frequency responses: one with a 4th order crossover and the other with a 2nd order crossover. Virtually identical transfer functions and completely different in character. Easy enough to understand, hard to explain and/or quantify as a whole.
Human aural perception, especially when properly trained, has incredible S/N discrimination as well as anomaly rejection capabilities. That is one of the reasons that intelligence agencies use computers to search for trends in communication and specially trained humans to hear through the noise quicker than they can do enhancement algorithms when time is of the essence.
A great experiment of the ability of the human to hear what machines can't is to set a mic up in a room and record a pin dropping to the floor. More often than not, the ear will detect it and the mic will not unless the gain is unbelieveably high and the sound of the pin is essentially lost in the noise floor of the equipment. Audio parlor tricks can be fun!

Dave