I once made on-axis measurements of several JBL 2440 drivers on a JBL 2350 radial horn and could measure no difference whatsoever whether the rear cover was on or off.
Newton's 3rd law would suggest that a broader range of parameters needs to be measured, no?

Certainly rear chamber loading/tuning affects excursion of the moving assembly.

In my own experience i've heard damping/stuffing in the rear chamber influence the timbral or sonic character to a degree as well.

I'm certainly no expert here, nor am i challenging Mr. Schell as i have a great deal of respect for his knowledge, experience, and contributions. Just trying to stir some discussion on a somewhat cloudy topic.