The discussion is going around in aimless circles.
(a) POH information on stall is for quasi steady state - slow reduction of speed (< 1kt/sec) to stall. Whether that be reasonable or not is not for me to say but, at least, it gives repeatable and quantifiable data of use to the certification process. If the aerodynamic warning is inadequate, one needs some artificial warning device.
(b) If you are yanking and banking the devices will respond to inertial and fluctuating airflow effects according to the design and installation. Why is anyone surprised in the slightest ? I'd be more concerned to keep a sensible lid on the yanking bit rather than worrying about spurious stall warning indications.
One can get an accelerated stall (increase/decrease) under g-loading (positive/negative). This should be covered in the principles of flight courses with the usual equation discussions.
.. as well as an unusual animal (usually described as a dynamic stall) where, under quite high pitch rates, the stall generates a short lived vortex which pushes the effective stall angle quite a bit higher than the usual item. Not a well scanned paper - the joys of 1-bit photos - but
this gives a bit of information for those interested.