For my money, your second point on "hydrodymanic forces" is a good avenue of inquiry. Such a twisting might be the result of a rotating body not hitting the surface in a more or less "flat" attitude (as did AF 447) but rather in a more or less "tilted" attitude - and it might not need to be rotating if it has sufficient ground speed vector to create torsion/shear.
Lonewolf, I started out with the aerodynamic. Then considered the hydrodynamic. But today officials stated that tail and FDR/CVR were 800 m apart, and tail and main fuselage plus wings 2,000 m. That increased the probability on the side of the aerodynamic. Also, because the vertical stabilizer still looked pretty smooth.
The damage to the lower side of the rudder area, that looks a bit like that on (high speed) AF447. Which puts another grain of salt on the aerodyn scale. Don't you think?