I believe the correct answer is to use whatever music you know and enjoy to test any stereo component. It's of little practical value to use bizarre, unfamiliar tracks for testing, unless that's what you are planning to play at home for your listening pleasure. Take a representative sample of your favorite music and you'll have a good idea of how any component compares to what you already know.

Hint: the best reviewers do exactly the same thing.
@spoutmouzert

Thanks for taking time to post all those "best" of links. It’ll be interesting to see what they include.

FWIW - Steely Dan's "Do It Again" is listed to test detail under the best tracks for testing treble category.

It changes as I get boarded with some songs hearing them over and over again.

Right now I use Dean Peer Travelogue and Think... it's all good CDs.

I like to hear how the bass guitar sounds and backing instruments to see if the tone is correct, the impact of the bass and how the backing instruments are layered and separated keeping the right decay and overall tone.  I don't want to hear bongos and drums sound like cardboard boxes but I want to hear the bong of the skin flexing.

IMO using poorly recorded CDs is just that a poor recording so I don't find that helpful in evaluating a system or component.

Happy Listening.