Maybe things aren't so difficult!
Using your example, if maximum rise @f22 for infinity subjects is 9.35mm then at 1:1 (double the bellows extension) the maximum rise is double too; namely 18.70mm. In-between values operate proportionally.
Mechanical vignetting happens when the exit pupil of the lens is cut by the format mask or lens barrel. The actual size of the exit pupil, on axis and off axis, depends on lens design but for a simple 47mm lens @f22 it is about 2.14mm diameter. For the same lens @f32 the exit pupil is about 1.47mm diameter. This means you can place the exit pupil closer to the edge of the format by 2.14mm minus 1.47mm equals 0.97mm. Not much of a gain in lens rise movement!
Optical vignetting is a non-linear artifact of lens design and it cannot, in practice, be calculated by regular trigonometry.
I have never seen any photographer use this sort of arithmetic in any picture taking strategy with a view camera except in a very approximate way. The ground glass (and film subsequently) is the absolute arbiter of image integrity.
If the arithmetic says the image is clear and the ground glass says it is vignetted then it is vignetted!
Bookmarks