Lyman, I think you are being fooled by asymmetric elevator deflection into perceiving the stabilizer is twisted.
On the first picture, you cannot make out any stabilizer landmarks, only elevator. On the second picture, the light gap between the stabilizer and wing is symmetric although extremely small. On the third and fourth frames, the light gap appears symmetric.
Keep in mind we are dealing with very few pixels, and if you blow the pictures up to gain accuracy, the surfaces become 'mushy' and ill-defined.
Furthermore the display appears to begin shading in the next frame over a short interval and you have the beginning of the next image ghosting next to the current ones. This ghost effect isn't all bad since the degree of displacement of the ghost is an indication of rate of position change.