the FDR readout is interesting, the deceleration rate was not rapid, nor was there a high sink rate for the step from F230 to F220. The automation was responding reasonably for the setup, just with a low target speed. The automatic entry from the angle of intercept will often result in bank angles above the normal value, to the 32 degree roll angle. (that angle can be exceeded, but it takes some effort from the crew to engibeer, and not in a good way). on level off, the pitch increase coincident with the bank angle and low speed starts the ball rolling.Aircraft get to this exact point every day around the world, the difference between a wild ride and routine ops being the initial speed.
In the level off and roll on hold entry, the speed tape min speeds would respond to the loading, but there is not much time available to intervene with something as simple as heading hold, thrust lever manual advancement.... to avoid the event. once it the event, the recovery procedure is in the QRH and FCTM. At high altitudes, excess thrust is limited, but at F220, MLW or below, there is copious thrust available.
Recovery involving a PIO is not surprising, during the event thrust changes are altering pitch moments and speed changes alter trim requirements, Unless hand flying of such cases are done in recurrent training, the PIO is not unexpected.
3 trained crew had a bad day, from an operator with a well established CRM/HF and cyclic training program, all speaking the same language... worthwhile lesson to them and any other operators of the potential to have a bad day.