jcjeant,
I can see why you infer that the BEA is contradicting itself, but that is not necessarily the case. As I explain in my post above. the THS reached 13 NU because the EFCS was trimming out a large amount of up-elevator. My GUESS is that the recorded 13NU represents the most that the auto-trim was capable of. Call it full-travel if you like.
Once the THS reached full-travel, the EFCS was probably having to apply more and more up-elevator to achieve what it wanted (i.e., what it thought the PF wanted). The brief nose-down inputs made by the PF may not have been sufficient to push the elevators into a down position. Unless and until that happened, the THS would remain where it was.