To calculate PA you have to know what the manufacturer means by his figures, and you have to trust him. Radpower and SixThreeZero have both burned me.
Radpower advertised their Radrunner at 750 watts and 80 Nm. It didn't meet expectations. To test torque, I found a paved driveway that would slow it to walking speed. With an inexpensive laser level I found the percent grade. I added my pounds to the bike pounds to get gross pounds. I divided that by 0.22 to get my gross weight in newtons. I multiplied gross newtons by percent grade and found that the tire needed to push me up the driveway with 180 Newtons.
The axle was 29 cm high. I multiplied 180 by .29 and got 52 newton meters, 65% of what Radpower advertised. To calculate power, I found a hill that it would climb on throttle at 11 mph. That was a good speed because it was a 20 mph hub motor, and I think power should be close to max in the top half of an e-bike motor's speed range. It was a good speed because in still air, 11 mph isn't fast enough for air drag to demand much of the motor. I calculated the newtons I needed on that grade, multiplied by 4.9 M/s (11 mph), and got something close to 500 watts, 66% of what Radpower advertised.
Bafang had certified the motor to output 750 watts indefinitely without overheating. That probably required a 24 amp controller for 1150 watts of input. For their demos, I believe Radpower used 35 amp controllers for 117 Nm of torque. The bikes they sold, though, seemed to have 16 amp controllers. They could claim 750 watts, but it would be input. I think they did this because they marketed the Radrunner for riding with a passenger. They realized the advertised 80 Nm would cause too many fatal back flips to sweep under the carpet. So they sold bikes with only half the torque they demonstrated.
With cadence sensing, ghost pedaling let me measure PA levels on that hill. A SixThreeZero Evryjourney produced 500 watts on level 5, as advertised, but on level 1, PA was 350 watts. You couldn't get much exercise that way because PA would take over. It was also dangerous. There were tight situations where I needed PA on account of steepness. That much acceleration would have crashed me. It seemed that SixThreeZero was assembling e-bikes from factory reject parts. I sent it back.
If you test at 17 mph, air drag is important. It depends on frontal area, coefficient of drag, and the cube of speed. A ballpark figure for someone upright at 20 mph is 265 watts. To find your drag at 17, you'd cube 17/20 (0.61)and multiply by 265 watts (162).
Such testing shows that the 750 watt motor on my 20 mph Aventon Abound puts about 930 watts on the tire.
