They were obviously aware they had a problem with the trim. I suppose that they didn't deal with it because they had a more serious problem at hand (UAS). They may have been reluctant to action the cutout switches without understanding the speed issue.
But, until the start of the descent, they were in control. I wonder what caused a change to their pitch control strategy.
A switch of PF/PM roles?
The red and green control force graph shows mostly the coupling with some limited red spikes. At the start of the descent, it's the green force that spikes.
Then, at the very end, the red prevails again.