group delay distortion aka envelope delay distortion. This is so hard to explain I am going to quote Dana Puopolo, who did a very good job in radio World article in 2005.
"To recreate any sound accurately, the reproduction equipment must have flat frequency response, low distortion and noise and a flat time response. In other words, the entire audio waveform must arrive at your ear clearly, at the right level and in the proper time. Group delay is exactly what it says: delay. As you approach the cut off frequency of a filter, the frequencies begin literally to slow down as they go through the filter. This means that they arrive after the fundamental and other harmonics. Problem is, humans can hear time delay distortions and filter group delay quite easily. We usually perceive group delay as a "phasiness" to the audio"Not all frequencies have this problem. Let me explain why. When you pluck the first guitar string it is tuned to "E" also known as 83 Hz. But it has harmonics. You hear natural harmonics at even intervals 165 Hz, 247Hz, 330 Hz, 659 Hz, 989 Hz, 1.3 kHz etc..(there are a mess of other harmonics at uneven intervals) This continues outward from the fundamental frequency beyond the limits of human hearing.
Radio wouldn't be interested in data above 20 kHz or below 20 Hz as those are the limits of human hearing. More subtly human hearing sensitivity isn't uniform, not between people, and not between frequencies. That is part of the reasons that the NRSC put the roll off at 10 kHz. That cut off means that only the frequencies above 7 kHz experience group delay distortion. But it does mean that people with sensitive hearing can hear upper harmonics above that frequency arrive after the fundamental.
In the book The Age Of Electronic Messages author By John G. Truxal rhetorically asks the question "Why did the United States adopt these regulations that doomed AM Radio to music of poor quality?" He then spent a few paragraphs describing what it would really take for an AM station to have high audio quality. I'll summarize. To reproduce the full range of a CD quality recording you would need a 36 kHz allocation. That would reproduce frequencies all the way up to 18,00 Hz. The problem is that then AM stations would need to be 36 kHz apart. It would require reducing the number of AM radio stations by about 66%. That really underlines the source of the problem.
The problem is human, not engineering. We have tried to find solutions in compromise. Physics isn't interested in placation. Reality is not negotiable. Listenership is already sliding. Trading around different painful compromises is no solution. The refusal to commit fully to one solution is the problem and always has been the problem. FM achieved ratings parity with AM in 1979. It's all been downhill for AM since then. My assessment is that it will continue to be the problem until no one is listening anymore. ...Or more cynically, until the remaining audience is old enough that they can't hear the problem anymore.