That's the same stick position as the spin recovery in a Tornado, and the aerodynamic logic in that case was to improve airflow over the fin when descending at very high AoA to try to stop the rotational motion. The stick was centralised once AoA came back on scale. A low probability of success was expected. I would hazard a guess that similar logic applies here.