And it's likely why the guy in the video had such a hard time with the bike in pedal-only mode.
I suspect the motor and bearings of the "wheel"

had a lot to do with that too.
If it's a direct drive motor, then at the large diameter there would be a lot of magnets, and a lot of stator teeth, and a lot of drag from just that.
Whether it is a single large bearing with needles or balls, or a bunch of small ones, that's a lot of bearing friction, too--especially if they greased them instead of a light oil coating.
I didn't watch the video: Did he ever show just hand spinning each wheel off ground, and the spindown time? If so, you can compare the front to the rear spindown time to figure out how much the motor and/or pedal-interface has to do with it, vs the bearings. If the time is the same, and short, the bearings are probably simply awful, which isn't unexpected. If the time is shorter for the rear, but stil short for both, then that's the motor / pedal interface *and* the bearings. If the front spins a long time then the bearings are not that bad and it's just the motor / pi.
The pedal interface should have some sort of freewheel or clutch in there so the wheel doesn't backdrive them and trigger the PAS, causing the motor to run by itself...but given the comments here I expect there either is no freewheel or it is greased so heavily that it is dragging the pedals along anyway, so it's also adding drag to the rear wheel.