What I find interesting is that if you look at FF, NL and NH in ENG2 they seem to be at Idle values (~200 and 60% respectively), not Off. And they stay that way for the entire period. Also the oil pressure never indicates problems.
It's Prop RPM that drastically reduces as the PROP is Feather to 86Degrees (or thereabouts).
All of them happen at the same time (within the same second I would say) so they don't look like manual ID and feather procedures on that engine (#2).
Could it be that something caused the prop to feather but not the actual turbine to fail? Any ATR/TP expert that could shed some light if this theory makes sense?