Reverse Engineering of the Power Meter

traditional cars are so boring: they just have an RPM meter that shows just the revs of the engine, or a speed meter that shows just the speed.
but what does an HSD power meter (CHG/ECO/PWR) really shows?
RPMs? Amperes? both of them?

put away your crosswords, this is a way more challenging puzzle.
can we desume its logic just by observing its behaviour?

beeing able to calculate the power meter from Torque PIDs, will allow to evaluate glide and other states of the car.

since we don't have any docs about it, we'll need a black-box approach: synthesizing inner working by observing inputs and output.

the output is well defined: the position of the indicator CHG/ECO/PWR.

the inputs are industrial secrets of toyota, I presume.
we may deduce them, and this is where the puzzle begins:
HV battery current is indeed considered.
RPM also.
and what else?

once identified x candidate inputs, the game goes on by defining f in:
y = f(x1,x2,...,xn)
f may be defined (polinomially?) by observing its values at given x values.

using torque, we may observe RPM and HV current values for every possible value of the power meter.
but it's hard to drive while looking at both torque and the power meter.

this is where crowdsourcing comes into play:
we may share observations between as many people as power meter readings.
every driver has only to concentrate on one single value.
after putting all observations together, we'll have a complete overview.

are you hybrid enough to accept the challenge?