As usual I seldom have any major problems with the qualitative
content of Gordon's reviews, but the final bullet points and numerical ratings drive me batty.
Note: the D600's factory-installed dust issue is worth mentioning, but assuming it's a QA issue that gets fixed shouldn't impact the final rating of the camera. (Any more than the light leak issue should affect the 5DIII or the left focus issue the D800.)
The 5DIII outscores the D600 by 1 point — OK I can buy that, the 5DIII has better hard controls, a better rear LCD, slightly better shot speed, deeper buffer, a better AF system, and more metal — but 3 points when you ignore "value" (which is a silly thing to have in the score but that's been discussed before). Among the "bad points" of the D600 are; no wifi or GPS built in (as opposed to the more expensive 5D3 which... also doesn't have them, or uncompressed video out) and 10.5MP in DX mode (vs. the 5D3 which doesn't have a DX mode because EF-S lenses block the mirror travel, and if it did it would have fewer MP in there).
The other bad points of the D600 are fair enough (exposure control during video and bracketing) but bear in mind that in objective tests (and no, not just dxomark) the D600 has 3EVs more dynamic range than the 5D3 (and there are some pretty impressive demonstrations of this
around). So three shots of bracketing on the D600 is roughly the same dynamic range as 5-7 frames on the 5D3. But hey, they have indistinguishable IQ according to the review (18/20). (I note the noise comparison between the 5D3 and D800 uses JPEGs for some reason.) Also note that dxomark finds the IQ of the D600 and D800 to be virtually identical, but the D800 gets 20/20 for image quality.
The simple thing is to ditch the darn numerical ratings, but people like 'em, I guess. If you're going to have them, they should be fairer. (And in this case, even the bullet points are unfair -- dinging the resolution of DX mode when the 5D3 won't even mount the lenses (and has lower resolution) is nutty.)