As they say, the proof is in the pudding ...
My highly un-scientific test seems to beat the scanner right off the bat.
Below are two uncorrected scans. The first was scanned with my 8800f at 2400 DPI in Vuescan, using multi-exposure (pretty laborious).
The second is with the K-x, JPEG straight out of the camera.
Even before any processing, the K-x image has more apparent sharpness and dynamic range. While the color balance is not as close to the real thing as the scanned slide, that should be easy to perfect with some tweaks.
How I photographed the slide: Placed my K-x with Macro-Takumar on a tripod, set ISO100, and stopped down the lens to f11. While I tried to position the tripod and camera at right angles to the slide, I really don't think it worked that well. Also, the film was not perfectly straight - I used the scanner's film holder to flatten it, but it still had some curl to it. The holder was placed directly on my flourescent lamp, the film was maybe 15mm above the light source. With some kind of diffusion, it promises to be even better.
Last edited by Dubesor; 02-18-2011 at 01:33 AM.