Originally Posted by jmastron

And it's not sufficient to model the MAX behavior with MCAS enabled and working; it must also model the behavior when MCAS fails, including whatever cues the pilots are supposed to have to diagnose (which ironically might be harder with MCAS range limited) and any control forces required to recover (e.g. how hard the manual stabilizer control might be to move), and the plane's non-MCAS behavior accurately (the extra lift resulting in inconsistent reactions to the stick).

In fact, I'd love to see every simulator (not just 737s) add in random MCAS-fail style intermittent trim until full down stabilizer events -- then we could get data to either confirm that every single "first world" pilot diagnoses and reacts immediately to this specific type of failure every time, or put to bed once and for all the claim that this is only an issue in the "3rd world". I know that's not feasible, but...