I don't think that the figures quoted apply to pressure per square foot, they apply to the linear force required to close the locking mechanism.
The locking pins were designed to baulk on a flange that rotated to clear their path as the locking motor drove the mechanism over centre. After the Windsor incident changes were made to increase the load required by increasing the pin length and then later by adding a support bracket that prevented the locking shaft from bending. The electrical cabling was beefed up too, presumably to try to reduce any voltage drop as the motor approached stall and the current increase then caused a torque reduction. These changes were supposed to put closing the handle on an improperly secured door beyond human strength, but had not all been applied to the THY aircraft involved.
The problem with the particular door on the Paris flight is that not only had the pins been shortened but the limit switch had been shimmed to indicate door closed when it wasn't anywhere near to being so. There was also a porthole with an indicator showing the position of the locking mechanism, however the operator working that day had not been trained on what it meant and also couldn't read the english text painted around it. Because of the mis-rigging the door closed easily under motor drive and the locking handle closed without any apparent effort.