Firstly, most probably don't know that Lloyd suffers from presbyopia (farsightedness); combine this with the normal age-related decline in vision by your mid-fifties, means that in poorer light, he needs to focus wide open to get accurate focus.
Secondly, there only is ever one plane in actual focus in a 3D image, the rest of the image is out of focus to some degree - the acceptable extent to which depends on how fussy the shooter is, how big you intend to print etc. So, having your main subject intersect with that single plane of actual focus can matter in large prints (simple being within the depth-of-field is not enough to render your subject the sharpest object in the frame), and focusing wide open increases your chances of nailing that focused plane right on your subject (assuming the lens has no focus shift when you then stop down to shooting aperture).
Thirdly, Lloyd also noticed that with the Voigt 35 APO and the Voigt 65 APO, there was some focus creep with the helicoid mechanism when the lenses were pointed downwards for shooting, which may also be 'contaminating' his results.
As to the Lenstip findings - that the 35 APO's performance dropped off faster at close focusing lab conditions at smaller apertures v's the other 35's, I wouldn't read too much into that: I spose its possible that the unique aperture blade mechanism could be worsening diffraction (because the circular-saw shape may increase the surface area of 'edge' the light has to diffract around), but I doubt you'd notice a difference in the real world.
My own copy of the Voigt 35 APO doesn't appear to have any of the above problems either, but I'm probably a little bit less fussy than Lloyd. I am, however, getting randomly overexposed shots: if for example, I focus at f2, then shoot an aperture sequence at whole f stops, touching only the lens to change the aperture, I'm getting an overexposed image - often at f8 or f11. The EXIF data shows the f top as being correct, but there is an anomaly in the shutter speed; i.e. on a cloudless day, you would expect the shutter speed to double for every drop in aperture, but it drops considerable more than that on a single frame within the sequence - and that image is overexposed: any ideas anyone?
