From that review, it is this page:
That is most interesting. Why?
Because about half way down, they take the same shot from the D800/5D3 and 100% crop of the shadows.
The image from the Canon sensor is as noisy as all hell (splotchy? banded?) whereas the Nikon one is noise free. That tells me that Canon still haven't fixed the noise issues that were integral with the 5D2.
All that post did was make me go test my lowly 5D2 the same way, push shadows the same degree, and conclude one of two things:
Either Lightroom totally blows chunks,
FM is not a site I will be visiting again due to lack of credibility.
Sorry, but recovering shadows with Capture One 6 gave me very pleasing results with 5D2 with none of the artifacts I see on the FM post. The D800 is better than what I get from my 5D2, but the difference is nothing even remotely like what is posted on that site.
Let me blow your mind a little. Converted using DPP and pushed in CS5.5
obviously these could not have been taken by a 5D3.............
http://www.flickr.com/photos/mattbicephotography/ here is my photostream so you can go look at the pictures at full size. I think LR4.1 RC certainly works better with Nikon.