And then there's the question of the different distortions added by the various microphones and recording techniques/machines used...
Of course Nrchy you are right in your critique, but there's still validity in the preference for acoustic instruments recorded live with minimal mic'ing in a real performance space, and with minimal alteration/processing during mixdown (or no mixing altogether). I won't go into all the possible reasons for this here, other than to say that it *is* very easy to hear for yourself that such records are usually far more revealing of the fine nuances of things like timbre and harmonic structure, low-level detail, and spatial cues than are recordings of electric instruments cut in multitrack studio settings. In other words, when I want to really hear what some change to my system is doing to the music, I always make sure to put on some more or less naturally-recorded acoustic material in addition to whatever else I might use, because quite simply it will show me the differences in the most revealing light possible.
None of which means that one can't or shouldn't use electric rock or whatever else to audition with, especially if that music accounts for most of what you listen to at home. Such recordings can and will be more demanding of systems in certain sonic areas where many acoustic performances tend not to be, and thus can reveal other qualities about a system. But what really helps is to keep the law of averages on your side, by auditioning many different kinds of music and recordings: If one system configuration can consistently give you a more natural impression - with more of your recordings - than some other system configuration can, than it's a good bet that system is, overall, providing reproduction that's closer to the 'truth', whatever that may be for any one given record. By the same token, utilizing a broad spectrum of recorded sources allows more accurate identification of any constant colorations or other distortions that are actually being contributed by the system itself. My own philosophy is of the 'master-tape' school of thought: The best you can strive for in a playback system is to 'accurately' portray what was laid down on the master-tape, as opposed to the live performance that tape was attempting to capture. Some will disagree, but I think that by trying to do the former, you are apt to get closer to achieving the latter on a more consistent basis. But you'll never get all the way there, or really even near enough for rock'n'roll...
One more thing: The very best, most revealing recorded source you can possibly use is the well-recorded speaking voice of someone you know well (not yourself! - we don't know what we sound like to others). But even this test only illuminates a relatively small portion of the frequency and dynamic ranges, and is not terribly complex or demanding a signal compared to group music. Its virtue is that we are intimately familiar with that voice as a reference, and that evolution has provided that our ears are most sensitive to minute variations in the human voice. The fact that this ostensibly simple test can never be satisfactorily met tells us all we need to know about what kind of chance we stand trying to reproduce live music convincingly.
Which is why, in my system, I don't try to. I just want my system to sound pretty good to me, meaning fairly uncorrupting of what my own sense about the 'correct' sound of decent recordings ought to be, with as little obvious sound contribution of its own as I can manage given my budget and commitment limitations. At the end of the day, if I can get close enough to where I'm not thinking about the sound of my system but just the music I'm playing, I really suspect that's all I can ask for. The funny thing is, this is very easy for me to do when I'm listening to a crap system, like the one I purposefully leave as stock in my car or my kitchen radio - I dial it in as well as I can, and then proceed to just get lost in the music, if it's good music. Only with the higher fidelity sound of the big rig do I tend to focus nearly as much on purely sonic attributes - is that self-defeating, or just the nature of the beast (or both)?