So what was important about those conversions? For my part it was to inspect the 100% image quality
without noise reduction . I know fairly well how other cameras perform when put in the same situation, so it is a very good way to gauge performance.
If you use a commercial raw converter like Capture NX or LR you
can't get away from noise reduction. Quite a lot of it is applied even though everything is set at "zero" in the user controls.
So, those images I linked show "the very worst" of the camera, at 100% magnification. That's the main point of the images, not artistic proficiency.
To get the images into a better perspective, here's the Canon 1D mkIV developed with the exact same amount of processing:
https://lh4.googleusercontent.com/-CrrlRrtcG8g/T05I20HO8cI/AAAAAAAAEPw/C43vTyGF9lU/s0/1D4_6400.jpg
Which is to be compared to the ISO6400 sample from the D800:
https://lh4.googleusercontent.com/-showvkoeOZ4/T0f6uSB37mI/AAAAAAAAENw/2ThCSeiLBbk/s0/06400.jpg
Hope that puts the images in a better context.
I never though the images would "propagate" outside the Swedish community without my original comments, that were paramount to understand the meaning of the comparison. My bad.