>The first of the 3 sets of pilots did this, and landed without incident.
Extra pilot in cockpit. Not a CRM situation that's a given.
>The next set overrode MCAS successfully 25 times, inexplicably did not turn off the trim, did not override it the last time, and crashed.
They didn't know why the trim was kicking in in the first place, and the senior pilot was the one countering the trim. He handed it off to the junior pilot to sift through the manual, which ironically, would have been futile anyway as MCAS was dropped from the manual. The behavior did not present as a classic stabilizer runaway. Other pilots interviewed also pointed this out including in the Australian 60 minutes expose on the 737 MAX if I recall correctly, so I'm not sure that the same definition you're fond of is universally shared amongst pilots. Bad priming is also a human factor that is catastrophically lacking among other things not being taken into account.
Furthermore, if pilots were really held to account for knowing the hardware they're operating, they should have known the nature of the trim switch change between the 737 NG and MAX, and the consequences thereof; that wasn't included.
>The last set successfully countered MCAS twice, then turned off the trim. They did not follow the directions on Boeing's Flight Crew Operations Manual Bulletin: "Uncommanded Nose-Down Stabilizer Trim Due to Erroneous Angle Of Attack During Manual Flight Only", which the FAA turned into a mandatory emergency airworthiness directive.
They did follow it actually, but also notice the Note toward the end (see the appendix in the Ethiopian preliminary report. The line reads)
>Note: Initially higher control forces may be needed to overcome any stabilizer nose down trim already applied. Electric stabilizer trim can be used to neutralize control column pitch before moving the STAB TRIM CUTOUT switches to CUTOUT.
Now. There is something to be said for "Gee, did they read it", but considering the bloody thing includes a disclaimer that the AFM supersedes the bulletin due to not being approved by the FAA, the note being presented outside the main content of the directive, and the notification beeping issued in November, I don't have high hopes for the real effectiveness of the bulletin at clearly communicating the severity of the problem, something which simulator time would have made plain and obvious.
Then again, that's just me.