Furthermore, and that is something you refuse to understand, the answer is implicit in the question asked: you are selecting the best picture a 35mm DSLR is able to produce and comparing that to the best picture a MF can produce and find out that they are of the same quality. Basically, out of 100 pictures taken by each camera, you compare the best one of each. You will indeed find that they are similar: one in a hundred selected here and one in a hundred selected there.
What your test does not show is that, for these 100 pictures, you will only find one or two at that level in the batch of 100 out of the 35mm DSLR but you will find 40-60 at that level in the batch of 100 out of the MF.
Last but not least, this discussion is only about the following criteria: sharpness, exposure and dynamic range, colour accuracy. May I ask why you are only interested in these aspects?
Jerome,
Don't let your passion get in the way of politeness! Civility has always been a feature of this forum, and it has served us well in the past, so let's continue that way.
We all encounter people here who we would not be meeting in real life, who do not belong to our own professional culture, and we need to find common ground with them. Cameras are designed and see first light in labs where no sun is ever seen and women wear bunny suits, and then they are exposed to supermodels and jungle climates
Now regarding the choice of images. My experience with MF has been that the keep rate is fairly low, because of technical issues with my pictures. My Mamiya shift lens was never really subjectively sharp, I don't know why, and it impacted my landscape images. Even with AF I had a lot of focus failures with the 80mm on portraits. The used 150mm I got cheaply was razor sharp, focused cleanly, and even took perfectly sharp images handheld under streetlights. Performance of the elements of the Mamiya system was unpredictable, and it is a complaint which I've found echoed in user-reviews on the web. I never found a fast portrait lens with an outstanding look. In the end I sold the MF system because the technical issues involved with bringing back a clean picture had become impossible to overcome consistently at my level of handheld incompetence. The keeper rate was often *zero*.
With the AF cameras I used, I found that the first images of a model session -often taken with an extreme tele in the street- were clean, sharp, and focused; so were the last, made in the studio. In between were some good and bad shots but the technically unusable were a small percentage. And of course in the end, the first and last images were usually the selects. Some of the Canon lenses eg the 135/2 have interesting "looks", I found for portrait. My Nikon files were consistently average with later models, focus mostly spot on, although the first Nikon I had, the D1x had very good color, and paradoxically I made a lot of nice portraits with the 17-35. In summary, I'd say that my chances of getting a usable retouchable first cut of the image I want with a dSLR, from a planned session, are just about 100%.
The Phase One had good skin tone, and superb color; the dSLR skin tones ranged from horrible orange peel to really good, with most in the "usable, nothing to write home about" category. Some of the dSLRs had good landscape color eg. 5D2, some unimpressive eg. D3x, with most somewhere in between.
At the huge shooting rate of dSLRs, and camera phones, one should look at the number of bursts, not frames, that yield a keeper.
One interesting camera I had was the Leica M8. This produced extraordinary images quite often, but had a bad habit of blanking out at random. Here I was virtually guaranteed several outstanding pictures per card.
The only camera I now use is a very battered Canon 1Ds3, acquired cheaply in a swap. It has a wonderul finder, fast AF and gives me decent colorful focused images about 90% of the time, decent skin tone, and is *completely predictable*.
If you have better success rates with MF and dSLR, well you are a better practitioner of MF than dSLR.
Edmund