If you are looking to evaluate how a system is performing in terms on timbre, imaging etc., recordings of live unamplified music are the only way to go, not because that type of music is in any way "better" that other types, but simply because it provides a fixed standard to judge from. An oboe needs to sound like an oboe and a cello like cello. And a symphony orchestra has a standard seating plan that can be used to evaluate accuracy of the sonic picture that a system presents. However, in the rock world so much electronic processing is done to the instruments that there is no standard from which we can judge the sound. In a modern studio rock recording the guitar sounds like whatever the artist or producer wanted it so sound like on that particular night. With no fix standard with which to compare the sound, we have a hard time knowing whether the system is doing a good or bad job in reproducing it. This doesn't make rock bad music, just a bad benchmark to do sonic evaluations from.
But while we can't really use rock to evaluate most of individual sonic characteristics of a system (I can be used effectively to listen for dynamics and system speed (slew rate)), I will not go so far as to say it shouldn't be in your kit of evaluation tools. We may not be able to use it to determine if a system sound "right" , but since many of us listen to a significant amount of rock music on our systems we should at least listen to make sure rock "sounds good" on them (of course "correct" and "good" don't necessarily go hand in hand). One thing that rock music is good for is judging the "goose bump/toe-tapping" effect. A good system should have the ability to draw you into the music and make you forget you are listening to a recording. It is always a good sign when you get so involved that you stop listening to the equipment and start listening to the music.A well done (from both a musical and a technical perspective)piece of rock played back on a good system will do this every time. I don't think rock could or should ever replace unamplified music as an evaluation benchmark, I think it can easily be used along side it.
A good craftsman's tool box contains many different tools, hammers, saws, screwdrivers wrenches, etc.. And while they can't be used interchangeably, each is very valuable within its give function.