The mitigating factor here is that closer to the spinner axis the backward movement is less than at the prop tip: I hear blade center of thrust was at 2/3rds the height, but max height is only in the center of the rotation in profile view, so the "scientific" way of doing this would be to calculate how high the ENTIRE 4000 lbs center of thrust is displaced into the upper disc half, as is accepted by PJGodzilla as being what actually happens on stick pull, and that height of the 4000 lbs center of thrust will then give us the leverage applied by the prop at full power against the wing's center of lift.
But then you have to take into account the leverage-increasing effect of the prop being at a right-angle to the lenght of the nose, creating a right-angle stress-riser that increase somewhat the actual leverage, given the forward direction of the thrust...
The FW-190A is advantaged, like all radials, by having less upper disc half leverage agains the wing's center of lift due to the shorter nose, and, by extension, less leverage compared to the elevator's tail lenght leverage...
Okay, I'll accept the scaling argument for now. Suffice it to say that the disk will resist rotation in pitch, both for the blade alpha and rotational inertial effects I've mentioned before.
Let's also be clear that you can resolve the resultant thrust + torque to a single force above the prop axis - thus admitting the change in center of thrust of which you speak. This is simply physical shorthand and is easily done.
The torque resulting from this CofT change, regardless of how far out you move it radially along the prop blades, acts on a pitch lever arm that is perpendicular to the long axis of the a/c - indeed should be something like the distance from the CofT to the a/c CG and will tend to pitch the a/c nose down. I see zero reason for the "close-couple" f-dub to realize any advantage or freedom from this effect that the Spit would not also have, this distance likely not being significantly different b/w the two. I.e., I think you're using the wrong moment arm in your mental model.
As for that "close- coupling" again, I'd say it's real advantage is a minimized pitch inertia. However, the ability to rotate the a/c in pitch is dependent on both this and the ability of both wing and elevator to generate a pitch moment - and we know the max lift of the wing per unit mass is compromised due to this a/c's higher W/A.
I don't think we get there from here with the 190.