In my experience, if you create a situation where Bubbah can grab a cord and pull it out when he is not supposed to, Bubbah will. If you train Bubbah not to, then Smitty will. You are setting this up for Murphy's Law to prevail. :happysad:
Also in my DIRECT experience with plug in connections and VFDs, the VFDs can strike and maintain an arc longer than standard AC, because remember, the output is actually DC pulses. SOME VFDs will detect the voltage rise of the arc and clamp or turn off the transistors, but some will not. It is not always part of the designer's thought process that anyone would think of doing something like that. If you ask any VFD mfr if this is OK, they will say No.
There are plenty of alternatives if you want a quick disconnect, but they usually involve pin and sleeve connectors, not twist lock. In that format, you can get them in a box where they are interlocked mechanically with a disconnect, so you cannot remove the plug unless the disconnect is open. I HIGHLY recommend that. There is also one pin and sleeve mfr who has theirs rated as disconnect switches by themselves(Meltric), so again, it disconnects power BEFORE you can remove the plug. Either of those is still not necessarily good for the drive, but they are safer for the humans. If you want to avoid stressing the drive, add an aux contact to the disconnect switch to turn the drive off before the main contacts open, then nothing is at risk.
By the way, unshielded flexible cord like SO cable on the output of a VFD? Really bad news. The output becomes a fairly powerful local radio transmitter. If you are going to do this, you should be using shielded cable, and ground that shield on BOTH ends.