Q: (Step 3) Removal of the original 'aperture limiting' cam. Is it necessary in order to re-install the bottom plate?
A: Not all lenses have this cam. If so, the bottom plate can be re-installed after the aperture-lever has been removed. In any case, I would remove "everything" to make place for the new baffle [step 5], whitch REALY makes an impact on the pictures. Test your self with and without the new baffle, and you'll see! Use small drops of superglue and matte black paper/ cardbord to attatch new baffle.
Q: (Step 5) New baffle: what hole diameter? does it depend on focal length?
A: The diameter of the hole is dependant on the focal lenght. 17mm diameter for the 40mm seemes OK and gives no corner shadows in pictures. Appr. 10mm for the 28mm. Mostly it's trial and error, guided with some knowledge of geometry (focal length, diameter of rear lens element and distance to "film") It would be nice to make a formula, given that the distance from baffle to film always is the same at infinity setting. Take pictures of a white wall at infinity to find optimal diameter. NOTE: With the 57mm 1.2 the baffle must be cut to make place for the contacts in the camera. I feared they would suffer too much strain and made the proper adjustments. No damage.
Q: I understand that e.g. a 40mm focal length lens will be 80mm '35mm equivalent' on a 4/3 system camera. Is there any additional 'crop factor' to be considered?
A: The 4/3 format makes more of the image circle than 3/2, so the hole in the baffle can be tighter. This is a good thing. How ever, some of the light-gathering power of the lens might be lost. Each lens should be customised to individual liking; sharper or softer, more saturated colors or not. The 57 1.2 is probably "masked down" to 1.8, but has not lost any of its strange bokeh. None of the Oly digi SLR's are designed for lenses with bigger image-circle than 25 mm, so the Konica-lenses give a fair amount of fogging on the picture if not "stopped down" a bit with baffle. I have not calculated the full impact of light loss.