Focus shift step choice is "interesting" to say the least. I find it safest to choose a small step size (e.g. 2) and then run a test set and then see what I got. But to figure this out may require some note-keeping if one wants to get better at making the appropriate guesses. On the positive, it does work well and with silent shutter it does it pretty quietly too, and without shaking. However I noticed with my 105 VR G Micro that the aperture doesn't always hold steady as the focus is being shifted, the lens compensates by adjusting apertures, but maybe in 1 or 2 out of 100 images, I got a clearly "wrong" exposure in the shot. It may be that it was -20C temperature, or my lens hasn't seen enough use (maybe once in the past 5 years).
I haven't noticed real-world significant exposure fluctuation in my other G lenses that I've used, but it is obvious that E lenses are more precise (when they work). Unfortunately I've had aperture total malfunction with E lenses many times over the years, in extreme cold. I have mixed feelings about current and recent aperture control systems in Nikons. Neither G nor E has worked with perfect results over the years. By contrast, when using lenses where the lens aperture ring was used to set the aperture and camera only "stops down to selected aperture in the lens" always worked well for me.
Anyway hopefully that incident with the 105 VR won't repeat as I start to use that lens more often due to it being my only auto focus shift compatible medium focal length macro lens. I'm looking forward to the next generation autofocus Micro Nikkors to better take advantage of the focus shift feature.
