what I would like to see is not a dyno run (unless its meant to improve power)

Go to a fuel station somewhere on the way out of town, fill it full tank of fuel. drive exactly 50km down the road and turn around and drive 50km back to the same fuel station.
fill it up again at the same pump, note how many litres it used.
connect the device up and take off exactly the same rate of acceleration, drive 50km down the road and back at exactly the same speed.
fill it up again at the same pump and note how many litres it used.
same day and within 1 hr or so means same temp, humidity and wind resistance.
same pump within 1 hr or so means same type of fuel ocatne, same error rate in litres.

that way u know exactly how many l/100km the same car uses with and without hydrgen input all else being equal.