To be fair, they should have tested the D800E rather than the D800 - after all, the H4D-40 doesn't have an AA filter either.
Dynamic range looks similar - Hasselblad has a bit more in the highlights, Nikon a bit more in the shadows. Solution? Expose the Nikon for the highlights, since the shadows can be recovered to a greater extent than the Hasselblad. And the DR values on DxO reflect the similarity in dynamic range between the two sensors.
The fact that Nikon persisted in using a 14-bit A/D converter rather than moving to a 16-bit version seems to have cost it in the tonality department. 14-bit may have been sufficient when CMOS sensors were still lagging CCDs in tonality and dynamic range, but, now that they have caught up, a 16-bit version might make more sense. Hopefully, Canon include a 16-bit converter in their next high-resolution body.
For landscape photography, however, I think the trump card is lens selection. Hasselblad doesn't have a real UWA, which is so commonly used in landscape photography, and its tilt-shift adaptor also acts as a 1.5x TC - which is fine if you have an ultra-sharp, wide lens and want to do some stitch some shifted images to really push the megapixel count (I wish Canon or Nikon produced a tilt-shift teleconverter, to stick on certain Zeiss lenses), but, without a wide enough lens in the first place, is of limited value when shooting landscapes.