A turn is not angled airflow. It is curving airflow. The airflow is slower on the inside turn part of the prop, which is not the case in a yaw (even for the split second the yaw is generated, because it could be that it is mostly the outside yaw prop portion that is accelerated, whereas in a turn, the inside prop portion must decelerate). This undoubtedly means a constant increase of the vacuum (in front of the blades) on the inside turn disc portion, which implies some kind of off-centered increase in trust that is nullified by the speed cost of the turn.
The reason you can get away with the notion of angled airflow on a wing (in a real curving turn) is because the height of the interaction is very narrow. This is far from the case with the height of a large prop, which is easily 10 to 14 times greater... Unlike a wing, any curvature implies a significant difference in airflow speed over such a broad height...
They cannot replicate this in a wind tunnnel, for obvious reasons... This is why I was asking what is the current consensus on this.
Gaston