No, that's exactly it. No recording sounds like the original event for many reasons -- primary among which is artistic license. We truly don't have a clue in most cases what the producer/recording engineer did in the process of creating a recording. Do you think jazz and rock recordings sound like the live event--esp. when reverb, delay, etc are added to the mix? Do you listen to an instruments two inches away? If you heard most of the cherished jazz recordings, they were done dry with reverb added later.
Now where do you think the comparison's begin? With us? No it's with the Producer trying to get the recording to sometimes more closely resemble the actual event and sometimes what the musicians desire. As Gary indicated, most engineers will tell you that you start out a recording with at least a 20% loss at the mike and then it goes downhill after that, especially if you're going through a console.
That's not saying it's desirable to have live music as a reference but the reality is that our frame of reference is actually what is put down on tape and hard disc.
Myles, when you put it that way, I would agree with you. However, i think we have to clarify one point....and that is, a recording that has been majorly modified to sound like what the engineer/producer thinks the 'live 'instrument sounds like and now in fact is a poor simile of same, isn't IMO a good recording of an instrument and therefore should NOT be relied upon as a reference. Even if one's system can totally reproduce the sound of said recording with no colorations or distortions of the recording, then while the system is great, the recording isn't. I would say in this instance, any reviewer or listener who believes that the recording is 'realistic' is fooling themselves and their readers. Instead, a comparison between what the 'live' instrument sounds like and the recorded version of same should be highlighted and the differences brought to light.That's IMHO.