Let me take a stab at your idea with my verbiage and new ideas...
Keep the basic architecture of the Andersen system, but change the ball/shank
As part of the ball/shank include a 'new' plate that is just below the ball.
That would then become the point of contact for the WD forces, which
would need a mating surfaces between them that would stand the
dynamic forces & pivot vertically to allow the WD to work.
That mating part could be part of another weakness solution of the Andersen system
The way Andersen clamps to the tongue is another weak design.
A cross member solves that. Have a longer than needed cross tube that is to
be sized and cut to length for each install. The clamps to the tongue now won't
have to hold it from sliding. As the cross member cut to length becomes an interference
fit wedging in the 'V' of the tongue
That would then have an extension up to the ball/shank to become the 'mating'
interface
The ball is literally just along for the ride in regards to the WD
forces. It then becomes a secondary or a safety for the connection
between the TV and TT
The ball shank bottom would remain the same
How does that sound?
The vector diagram needs to be generated, but this first blush of
a modification to the Andersen design looks good...but a bit of NIH
here as usual... :)