IMO pop songs are more useful for evaluation. Because they normally include multiple tonal and transient elements, some very unnatural elements, some natural but heavily processed, there's only so many ways to reproduce coherently.
With real acoustic events like an orchestra, we can't know with certainty how each instrument sounded in that particular room. It's placement in the room, the presence of other musicians or and audience, all interact with the sound of the instrument. At the engineers discretion the event is captured within the constraints of the recording equipment, and mixed within the constraints of a stereo mix, a medium that forces you to take much artistic liberty just for practical reasons. In many ways an orchestral recording is very different from hearing a live acoustic event. That makes the idea of judging an IEM by your memory of an instruments sound quite nonsensical.