No, I agree with Jerg.
For instance, frequency response measurement is typically done by taking into account the highest peaks of impulses sent by a frequency sweep. The highest peak is one thing, but looking at an RTA read out, I can see there are other less high peaks that resonate out at other frequencies due to various factors (enclosure reflection, ear pads?) and those peaks are definitely audible, too. I suspect they are related to harmonic distortion, or something along that line.
I have tried to eliminate those extra peaks just to see what happens, and sure enough... extra resolution, and somewhat faster decay, but the frequency response sweep still won't show this, and I suspect CSD won't, either, since CSD deals mostly with decay of a particular frequency rather than as a whole (but please correct me if I'm wrong!).
I really need to put together my own measurement rig one of these days and conduct more thorough studies...