I wonder how long between selecting flaps and disengaging autopilot?
It matters not the aircraft type, if you change the configuration and follow it rather quickly with an AP disconnect before the auto-trim has finished its job you can be guaranteed a surprise... A pitch up after flap extension (especially if it was on the upper end of the flap speed envelope) would have to be expected, surely? It does sound like it could have been purely normal physics combined with poor airmanship/timing to cause an event like this. I have flown aircraft where the AP can spit the dummy and take longer to disconnect when interrupted while auto-trimming. It's a bad time to interrupt an AP!
If it was a true trim runaway one would think the only thing that would isolate the situation would be to remove power from the pitch trim mechanism through a 'trim interrupt' type switch or by opening a CB. There doesn't seem to be any info about such action which would seem to suggest that after control was ultimately regained from the AP the trim worked normally?