In theory, they should be about the same. In practice, the Canon is better engineered for this kind of scenario and so will produce a much better image is my guess...
It's my guess also, but we're now in the digital age. There should be no need for guessing. It's now so easy to do comparisons under any conditions one wishes to specify.
I recall that illuminating comarison between 35mm film and 4x5" film, produced many years ago by Lars Kjellberg of Photodo. Do you remember that?
As I recall, the films used were the B&W T-Max 100 for the 35mm camera and T-Max 400 for the 4x5 format in order to at least increase the shutter speed a bit, at equivalent DoF.
The object of the exercise was to see how 35mm compares with 4x5 at equal DoF, without resorting to the use of tilt lenses.
The 35mm camera was used with an exceptionally fine lens at F5.6 (might have been a Leica Summicron, from memory) and the lens of equivalent focal length on the 4x5 camera was set at F22, a difference of 4 f stops, which matches exactly the ratio of the differences in the height of the two film formats, in order to match the DoF characteristics.
The results were surprising. The 35mm shot was actually slightly sharper than the 4x5 image. However, the 4x5 image was less grainy. With today's noise reduction programs, I guess one could sacrifice the slight edge in sharpness of the 35mm shot to get a result very similar to the 4x5 shot. Except of course, the faster film of ISO 400 was not sufficient to enable use of the same shuter speed with the 4x5 format. To do that, one would have needed to push-process the T-Max 400 film to ISO 1600, resulting in a 4x5 image which I guess might then have been noticeably inferior to the 35mm shot.
So the simplest answer is: for any shots you take below f11, in theory those could be sharper with medium format digital (and yes, f11 is a lot sharper on medium format than on full frame, just as f11 is sharper on full-frame than on APS-C). But of course you would need WAY more light and better lenses.
Aren't all lenses equally bad at F22?
My own tests involving different formats (mainly cropped-format DSLRs and full frame DSLRs) is that F22 seems to be the F stop where the lens is fully diffraction limited. I noticed with my Canon 5D, which has bigger pixels (or at least bigger photosites) than the P65+, I could use F16 without any major concern about resolution loss. The resolution differences between F8 and F11, and between F11 and F16, were difficult to see even when pixel-peeping at 100%. However, the differences between F16 and F22 were quite obvious, F22 being quite noticeably softer.
I imagine that users of MFDBs would want to stay clear of F22 if sharp images are their goal, especially considering that lack of an AA filter. At F22, no camera needs an AA filter. It's no wonder that Michael used F11 when he compared a P45+ image with the P&S G10 at F3.5, at A3+ print size. To equalise DoF he probably should have used F22 with the P45+, but that
might have risked the ludicrous situation of the G10 out-resolving the P65+. I say
might because of course I'm guessing again.
I should also mention that I've compared the resolution of my 10mp Canon 40D with the 15mp 50D at various F stops, just to see what effect diffraction would have in limiting the potential resolution increase of the 50D.
Once again, F22 appeared to be the F stop at which no resolution benefit of the 50D was noticeable. At F16 the 50D delivered about the same resolution as the 40D at F11. The 50D also delivered about the same resolution at F11 as the 40D at F8. However, at F8 the 50D was slightly sharper than the 40D at F5.6, but that was probably due to the quality of the lens. I was using a standard Canon 50/1.4 which is probably only very slightly sharper at F5.6 than at F8, at least my copy is.
Medium format digital capture emulates the workflow of your RB67...it assumes you want to use low ISOs (like with film) because you'll be shooting landscapes or in studio with strobes (and always with a tripod). The dSLR is much more general purpose. The article is just trying to argue that medium format digital is a little more flexible than it seems, even if it's not designed to be very flexible.
Fair enough! MFDB gear can only come down in price if more people buy the equipment. Perhaps I should 'pull my punches' more.
But I still think, if one values flexibility in a camera system, James Martin's article could be misleading for some potential MFDB customers. Sometimes we
want something badly, but perhaps we realise it's not quite pratical for our needs, but we rationalise to ourselves, and kid ourselves, that the disadvantages are not serious.
I sometimes can't help sympathising with Jeff Schewe carrying 35 LBs of P65+ camera gear and having to leave his 1Ds3 & lenses behind. I certainly would, in his circumstances. Even 35 LBs is too heavy for me.
The last time I was in Nepal, photographing the spectacular Himalayas, I carried two cameras, the Canon 20D and 5D, plus a Sigma 15-30 zoom, Canon 24-105 zoom, Canon 100-400 zoom and lightweight Manfrotto tripod. Total weight would have been about 15 LBs or 6 Kgs, for a continuous focal range of 15mm to 640mm.
That's what I call flexibility.