Can't say I've noticed congestion so much -- although I tend to listen to simpler stuff than, e.g., orchestral music. They still strike me as tonally not quite right. There"s just something un-natural about them. Not euphonically un-natural (plenty of that in "hifi" now and in the past!)... just unrealistic. Not sure what the root cause of that is, based on the quantitative assessment.
Seems like the quantitative testing is mainly for tone, dispersion and distortion. However, there isn't a test for how close to real the reproduction is. Like how real does it reproduce the pluck of a string, a tap of a cymbal, or a vocal? In the rest of the testing world there are usually reference standards that other instruments can be calibrated against. The problem is no one speaker can do everything well so it seems like there needs to be a set of reference standards for audio speakers. So there would be a reference setting committee that would select one speaker to serve as the reference standard for, say, true and real reproduction of a female vocal. In this example, this female singer would be present in the room and the reference standard chosen would have the most realistic reproduction of that singer to actually being present in the room to hear her.
Of course, that reference committee would have to select all the components that come before the speaker to record the original sound in the truest way possible.
Perhaps there would be 10 categories of basic sounds so there would be a kit that speaker designers and manufacturers would get to compare their speakers against the reference standards for the realistic reproduction of musical sounds.