The voltage spikes developed by the inductance of the motor windings can destroy the semiconductors in the VFDs power bridge.
I am having a hard time with that.
How is a spike from the motor that was caused by the switch opening getting back to the motor through an open switch?
I do agree that opening a switch on the load side of a VFD is not a great idea but I do not think that this spike is the reason.
That said it is common to have disconnect switches on the load side of VFDs when the VFDs are located at a central MCC and the motors are not in sight.
As 'Frenchelectrician' mentioned installing an auxiliary contact in the remote disconnect to disable the VFD is common.