Subjectively speaking, most of our hearing can’t be measured in a way that tells us how something will sound. So if we can’t measure our hearing to correlate with the numbers ( measurements) used by manufacturer’s of audio components, why are they so important? Assuming I’m right?Actually we know quite a lot about how human hearing works, but a lot of it has only be discerned in the last 40 years or so. The spec sheets are based on tradition and are taking a while to catch up.
So if the designer is aware of research into human hearing perceptual rules its not that hard to take those rules into consideration when working on a specific circuit- and it is possible that such a circuit might not measure all that well (which usually means that harmonic distortion might be fairly high) based on what is considered 'good' by tradition.
Here is an example- the ear is relatively insensitive to lower ordered harmonic distortion (2nd, 3rd and 4th) but is keenly sensitive to the higher ordered harmonics. So if an amplifier has fairly high 2nd harmonic, but is very low in the higher orders, it will sound smooth and easy going to the human ear. If the amp is low in the lower orders, and also low in the higher orders such that it might have only 0.01% of THD, its very likely that it will sound harsh, as that is the property that the ear assigns to higher ordered harmonics. You really have to get those harmonics much lower in order for them to be considered 'inaudible'- 0.001 **might** be acceptable if 'bright' and 'harsh' are the subjective experience that you are trying to avoid. But it seems that you have to be even lower than that.
Now there is an old saw that if you have high THD you will also have high IMD (intermodulation distortion) and that is often true, but not always! The ear is sensitive to IMD and it is arguable that if getting low distortion is a goal, that keeping IMD down is really where you want to concentrate your efforts.
So this is sort of the tip of the iceberg about why I feel that the quote above is a false assumption.