Lighty,
you are probably right about those various factors playing tricks on me
I played around some more and noticed that the MOSFET that is getting cold to the touch is that on who is currently not working, I mean, if I oscillate phase1 then the MOSFET of phase2 gets cold to the touch. It is possible that the metal part of the MOSFET just feels cold but in fact is at room temperature, like all metals feel cold to touch even at room temperature.
I will make these RMS measurements you said

This is what I learned today.
I connected the output of the motor to a battery. I put an amp meter on the positive side and the scope leads across the amp meter to see the output current waveform. Then I adjusted the duty cycle of the pulses and observed how the waveform changes. As I noticed before, there is a sweetspot in duty cycle where it is possible to get the best recovery and it is somewhere at 56-57% positive duty cycle.



Here we see the output current waveform with less than 50% ON time. You can see how small the waveform is in amplitude compared to next waveforms.



This is the same waveform, only this time the duty cycle is at about 56%. You can see how much bigger in amplitude the waveform is.



And this waveform is with 60% or more ON time. We can see that the off time becomes too short for the output current to discharge properly. With such duty cycles I can get only 50% or less recovery.
Thanks,
Jetijs