Detecting Propeller load variations on the vertical plane
What I am interested in is a question of basic prop physics, for a configuration like that of WWII fighter types: How would a vertical (to fuselage) prop load variation be detected, since, on take-off, any vertical tendency would be held by the ground one way, and the airplane's weight the other? The exact same holds true in straight level flight, or vertical pull-outs: Wing lift being then substituted to the "ground": These much larger vertical forces would "hide" any vertical variations in the propeller load...
Maybe steeply banked horizontal turns would be more sensitive to this vertical prop load variation, since some types are known to require stick pushing while turning at lower speeds, to maintain the turn, which indicates some kind of anomaly in the vertical trim of the aircraft. One that seems only detectable in sustainable speed (lower speed) steeply banked turns...: Interestingly, lower speed turning is where wing lift is weaker, but propeller load is at its highest... This effect usually disappears at higher speeds.
What I am specifically interested in, related to this, is what is the current flight physics consensus on the effect of the wing's presence in altering the uniformity of the vertical loads on the propeller. What I mean is that, relative to the turn's curved airflow, the wing in effect drops in relation to the prop disc in a turn: Would this not disrupt, vertically, the uniformity of the loads inside the propeller?
The reason I am asking is that all of the well-known prop load variations (P-factor, torque, splistream spiral on tailplanes) are either lateral or rotational in nature, but purely vertical prop load variations would be far more difficult to detect, since the ground or wing lift is "holding" them one way, and the aircraft weight the other...
Exactly what is known about the wing position's effects in disrupting the vertical uniformity of the prop's thrust, since even large effects would be completely undetectable to the pilot, except maybe as raw comparative turn rate data?
Gaston