About the sidesticks. It is my understanding that Airbus sidestick inputs are taken as requested deviations in control surface positions, and not as states for these positions. That is, when a pilot wants to adjust the attitude of the plane by some finite amount, s/he pulls the SS back/forward a specified amount, AND THEN RELEASES THE SS which then returns to a neutral position. The amount of the initial deflection in the SS is what determines the degree of attitude change. Holding the SS at a deflected position causes continually cumulative control surface inputs in that direction.
Thus, isn't it extraordinary that Bonin kept the SS in a deflected back position? Or is my understanding of how the SS works incorrect?