There was a recent thread on the Gon that addressed a few of these issues: accurate to what? We listen to recorded music over our systems. Either it is studio recorded (with all the artificiality that involves) or a recording of a live event (which, as you pointed out in an earlier thread, Tim, may also involve some sonic manipulation, perhaps less so). None of us (or few of us) have had the opportunity to hear the studio performances as they took place (if they even took place in the same place at the same time). And, for 'live' recordings, unless you were at the event, and have not only a good memory of the sound, but can also take account of where you were sitting (since that's going to make a difference too), we are all shooting in the dark a bit on whether a hi-fi system replicates with verisimilitude (love that word) the original performance.
With respect to just measuring the playback equipment itself for 'accuracy,' i think even the objectivists agree that the measurements only tell us part of the picture, and also don't take account of real 'room' conditions in playback, as well as 'synergies' among components, including cable, as part of a 'system.'
So, what's the basis for evaluation? For me, it's knowing what a real trombone or cello sounds like. Not the cello on that recording (although i may be able to distinguish that one, from experience, from a different cello played on a different recording and might have even heard that particular cello, played by that cellist in a live performance at some point, but look at all those variables, even with the same instrument and artist). Does the playback have (enough of) the sonic attributes of a real instrument playing so that I can achieve the illusion of a real performance in my room? (Obviously, a noisy recording or a bad one will make it harder to suspend disbelief, but using the best possible source material, does it sound like a real instrument or voice, recognizing that there's no way my A x B x C listening room is going to approximate the sound of a big hall?). Yes, it is subjective, but I don't think its entirely dependent on 'what i like is right.' One way to tell- and this was part of that thread I mentioned elsewhere- is that different records are going to sound different; they are not all going to be colored in the same way because the system is tilted a certain way, sonically. I can't take credit for the last observation, but it all fits. And, i do think, in this context, there is a role for 'experienced ears,' someone who has been to a lot of performances, knows what live music sounds like and can make judgments based on that experience.