It is worth noting however that a dust particle that is sitting above the film allows diffused light to pass under it from the side (both because of its shape and because of the thickness of the film base), hence reducing it's ability to block light and therefore cast a shadow on the paper. The image forming silver however is in the emulsion, and far "shorter" than a dust particle, so non-perpendicular light will have a lesser effect. The idea that diffused light sources make dust less apparent, so therefore must cause loss of image sharpness is I think, flawed because of this.
How much effect diffused light does have, I have no idea: I'm sure there must be some, but how much, is open to debate. The fact that some see a difference and others do not suggests that any difference is slight. No doubt enlargement ratio is also a factor with small format printers seeing a greater difference because of the greater multiplication factor when going from a small negative to a large print.
I wonder if anyone has done a serious study of these effects?
Cheers, Bob.