Too little information at this time.
There is a lot of important information missing from the plot as well.
Autofeather on the ATR used to be linked to the thrust selection knob whose position is not reported. Moving to climb would cancel uptrim and autofeather. Depending on what profile they flew this might have already been the case.
Did a few hours on type ATR 72-200's and it should be controllable with a non-feathering prop at any speed above Vmca. Be quick but you should have the "one" second to react and make sure you got it right. I can confirm the torque increase when going to feather making a bit of a yaw to the good side before telling you clearly which one has failed. Just keep your nose straight.
We'll have to see what the DVDR and CVR analysis together with the data that can be recovered from the ECU's and maintenance computer will show once examined.
IMHO the funny things in the data plot released are:
Why did they move the PL's so erratically?
Why the funny movement of the PL #1 in several steps to idle?
Why moving PLA #2 to MAX shortly before 02:53.15
The fuel flow on #1 increases shortly when #2 fails and nearly at the same time the PL #1 is reduced.
The PL #2 stays at TO and is then increased to MAX.
The rhythmic fluctuations in torque and oil pressure need an explanation.
What exactly is TQ1 and TQ2 versus Objective TQ1/2?