Originally Posted by IF789
I think the risks either way are going to be very very difficult to quantify which is why this is a tough call.
That is what the engineers get their pay for...making tough decisions on system design.
I don't design circuits for a living or write software, but I've done a fair amount of root cause analysis on defective equipment and have had pretty good exposure to same over the last 55 years. The following is a gut feeling hypothesis only so take it with a healthy dose of skepticism:
Once all 3 airspeeds become corrupted from high AOA, if any part of the primary or supporting calculations for operation of the trim rely on airspeed or its derivatives, then operation of the trim becomes compromised. Without looking at an actual complete logic diagram of how the system operates, this is difficult to analyze.. Something as mundane as a calculation of how much torque the hydraulic drive motors must generate to drive the THS (just for example) can have airspeed components in it that crash the whole logic chain for THS drive. Any reasonable software would then have error handling options for such an error. Generally the safest option would be to halt all automatic motion on error.
Now I could be seriously wrong here, but
something stopped the THS from trimming nose up before it actually hit the end limits. If it was an error handling routine, it might stop nose down trim as well until some reliable airspeed information appeared. In this case you might effectively be in "Direct Law" without an indication, at least as far as THS trim goes. Elevator motion would then still be Alt2 though.