This may be a load of bunkum, but for what it's worth –
I don't think this is earth leakage as in insulation resistance at all. I think this is about circulating currents, though I don't yet see the mechanism for an earth leak trip to pop out.
In the days before the idea of VFD's was ever thought of, motor designers didn't need to take into account the effects of high frequency components of the applied voltage. So some early motors (as I have been told), when fed from a VFD suffer from the induction effects of high frequency current internally causing all sorts of strange effects, usually associated with heating effects, bearing wear (likely because of currents induced in the bearings) and general inefficiencies of all kinds. Not all three phase motors are compatible with use on a VFD.
Motors of a more modern construction are better able to cope with the consequences of having edgy waveforms superimposed on the supply voltage and behave as one would expect.
Taking this to more modern again, the cost of fancy semiconductors suitable for mains voltages has reduced significantly, and VFD designs use this to reduce the obnoxiousness of the AC waveform applied to the motor, getting better approximations to a sine wave.
I suspect your audit of motors has thrown up a cross section of the historic timeline of the steady improvements in motor design.