Obtaining linear response by stylus and cantilever waving shadow plates between LEDs and optical sensors seems problematic;what do you make of the idea of an optical cartridge - it's likely advantages or disadvantages?
optics for consistent beam shaping, diffraction, wave plate resonances want considerable sorting.
Optical beams tend to be round, with vignette gradients; passing an opaque straight edge stop across such beams will change photon rates
with a sinusoidal correlation to radial movements, to a first approximation. I suppose that each cartridge can be individually calibrated,
but wonder about stability with e.g. suspension aging and how airborne contamination accumulation in the optical paths is compensated.
My understanding is that good blue LEDs are easier than green.The invention of the blue LED changed the entire world.