Could be for intentional slippage (to reduce the paper "damage" or extent from jams and thus make clearing them easier...?!).noted that voltage is proportional for how well gear grips to hub, which makes sense, so it could also be used as linear-clutch, instead of on/off clutch.
Yes, this is what it does.Reading, or in interpreting the Patent description, it appears that input and output shaft are separated until the coil is energised, whereupon this results in winding the spring up after which the secondary shaft turns.
It matters somewhat in a clutch used for tensioning. You generally want to go negative a bit to counter-act the residual magnetism if you want close to zero. I built a tensioning controller and didn't do that, but it turned out we didn't need it anyway.Correct, usually polarity does not matter unless a motorized device OR a magnetic latching solenoid, then polarity matters.