I just did a slightly different test, and was surprised how quickly the difference became apparent. I tested 4x5 vs. a 6 mp DSLR (Nikon D70), which was obviously going to be a big difference at huge sizes. My test was a LOT less scientific, but was perhaps representative of how people really use the cameras. I did not use the same shot, but picked one from my first day with 4x5 (which I'm just learning) and one from my extensive file of D70 images that didn't show any of the downsides of digital (base ISO, no huge dynamic range, etc...) I scanned the 4x5 Velvia on an older Epson Perfection 3200, at only 1600 dpi (I know that's a 3200x6400 dpi scanner, but, as has been discussed here before, it can't really capture its full resolution).
I did two print tests, and judged on the results from the prints, not from huge pixel blowups (the pixel blowup would be NO surprise-4x5 just beat a 1Ds II, a D70 isn't going to catch it.) First was a 16x20 print from 4x5 against a (roughly) 14x20 print from the D70. I matched the 20 inch dimension and let the width fall where it may. These prints were produced on an HP DesignJet 130. The D70 print looks OK viewed alone, but it is best not to view it with a print from 4x5 in the same room! The detail in the 16x20 from 4x5 was simply amazing...I was not surprised, except, perhaps, by the magnitude of the difference.
After the 16x20 result was so conclusive (I consider the D70 to be"falling apart" at 16x20), I decided to see how small I could go and still see a difference. 8x10 (and only 6.7whatever in width on the D70) should be closer? Time to fire up the Photosmart 8450-a very modern, high resolution, wide color gamut small printer (the only printer thus far to have a native color space of Adobe RGB). At 8x10, the D70 is a LOT more acceptable than at 16x20. However, the 4x5 is still easy to pick out-the 8450's pulling detail from the scan that just is not there in the digital file. The 8450 is so high resolution that most of the detail from the 16x20 is still there-you just have to look closely. There is a depth to the 4x5 print that is simply not present in the digital, even at a small size. I wonder if it is better dynamic range (although the 4x5 was Velvia and the digital shot was chosen not to push the limits), just the increased resolution, or what?
I WAS surprised to have a visible difference as far down as 8x10! I thought that a good DSLR operated within its limits was essentially perfect at a print size that small (I thought that the only reason to use the view camera for prints that small was for the movements). It IS very good, but I can pick out the view camera every time even at 8x10...
The view camera in this comparison was a pretty simple rig-a Wista DX wooden 4x5 field camera with a multicoated (but fairly low-end) Caltar (Rodenstock) 150 mm f5.6 lens. Film was a Velvia quickload in a Polaroid 545 holder (supposedly not the sharpest way of doing things, but plenty to beat digital).
-dan