Take a picture of your DMM, and I'll tell you how to set it up for current readings.
So far, we've established that 100% power & 75% power gives the same resistance. OK what? 75% power should give 25% resistance. 0% power (off) should give 100% resistance. 50% power gives 50% resistance. The more resistance (higher Ohms reading) the less power to the grips, so the cooler they are.
I would expect that the new style controller is still just a variable resistor (with a bit of extra trickery for the battery level detection). If it was simply a resistor, the above paragraph would stand.
If it's also a pulse generator (power turned on & off quickly by the controller, timing depends on power setting), that might account for the fluctuations in readings. But I doubt they'd go to the expense of a 12v DC pulse generator just for that unit. Seems very overly complicated to me if they have, and therefore more prone to failure.
The voltage jumping around seems to indicate the previous paragraph (it seems to show that it's a pulse generator AND a resistor).
To measure current, you need to be in series with the grips. To do this, 2 small croc clips & a bit of wire. Connect one pin from the controller using the croc clip, then connect the other end of the wire to the matching pin on the grips connector block. The other wire gap is filled by the DMM.
Basically, to measure voltage or resistance, you can be in series or parallel, but to measure current, you have to break the circuit & fill the gap with the DMM.
If you're still running around in circles mate, I'm free this weekend & can meet you somewhere like Bala/Dolgellau with my DMM
You could even take my grips off at the same time
To be honest, it looks like the controller is at fault. Are you doing these tests with one grip plugged in, and the DMM where the other grip should be?