I think the key piece of missing information here to make this more understandable is that ground is often not a good conductor. As a result you can have different ground voltage levels even within a distance of 100 feet.
When you ground both ends of a shield or any other kind of good conductor, you form a good conductor for ground currents (most of them probably unrelated to the signal you are working on). These currents can be surprisingly high and can generate unwanted fields in the shield thus inducing unwanted signals and noise into your signal wires.
QUESTION OK so the lood and resulting current in a shield when grounded at both ends can be a result of potential diffference between the ground points. That is most illuminating.
In the case of drive/motor leads, the shield is made to carry high currents. It is also important for the motor frame and the drive frame to be at the same ground potential in order to avoid nuisance drive faults. The motor leads are already full of high frequency components so a little induced noise from the shield is hardly any concern. And the motor is not particularly sensitive to noise anyway since it is primarily a large inductor.
QUESTION This is the reason I would take it that a VFD and all associated "pieces" should have only one ground point.
Just as a footnote to this. I have seen ground potentials between adjacent buildings in an industrial site with over 30 volts difference in potential. Attempting to ground the two together in one instance involved over 400 amps of ground current, most of it at 60Hz, of course.