I know that you (and others) believe that as displacement goes up so does "distortion" and hence lower sound quality, but there is no evidence that this is correct, and I highly doubt that it is.
I assume you're familiar with BL(x) and Cms(x) ?
Have you ever modeled what this does to the actual system response?
The BL in particular.
There is nothing left of that system response anymore.
I have seen cases where the Fs shifted extremely significantly (well over 30-40%) and the Qt (= BL) peaked well over 3dB.
And before you're gonna say the same thing again, yes from actually high quality expensive loudspeakers.
And even in a situation with room modes, these effects are noticeable in a relative (quasi) static environment.
Any music noob can hear these kind of changes in EQ.
Sure this isn't a problem when big woofers are being used at baby sound pressure levels.
(since there is barely any cone excursion)
But since you put this as a general statement without additional context, it's important to explain the nuances.
Also, there is a direct relationship between Le(x) and (intermodulation) distortion.
I don't see much relationship between Cms(x) and any intermodulation components.
Although in theory this will always change IMD components, since it changes the system response itself as well as the low frequencies complex harmonics.
But a dropping/changing BL definitely has a direct effect on the overall driver performance as well as back EMF components.
I might be mistaken, but I think Klippel has a paper about this.
As for that distortion is always "bad", that is a totally different can of worms buried in a deep rabbit hole part of a philosophical chicken-egg story.