The problem with the test is that the most awaited question, the quality of the AF, which has been so seriously lacking up to a least the K-5, was answered rather subjectively.
For me, who tries to shoot a lot of sports with the K-5, it's the AF.C performance that I really want to know about. The adjustment speed of the Sigma 70-200/f2.8 HSM should be sufficient to reduce the factor of lens AF speed from the mix, allowing more attention to be focused on the camera's AF.C tracking performance. I agree with falconeye that the logical comparison in the Nikon range would be a D7100, not the D610, both with Sigma 70-200/F2.8 HSM lenses.
It would be nice to have standardised AF.C tests. This would require shooting from the same location, with equivalent lenses & settings, of say a car coming at a certain angle and speed towards-across the camera. Ideally, the lighting needs to be similar to ensure repeatability, if not from day to day, at least on the same day. This test is done by a French site (
http://multimedia.fnac.com/multimedia/editorial/labo/reflex_2012.pdf,
http://www.fnac.com/Pentax-K-5-II-Noir-Obj-Pentax-Smc-DA-18-55-mm-f-3-5-5-6-...6/w-4#labofnac), but not widely known outside Europe.
Another thing is subject colour. Tests could include similarly- and a differently-coloured subject to the background to determine how well colour is being used to maintain the lock on the subject.
Also a similarly-coloured nearer, unwanted subject, e.g. emulating the situation of two players in similar jerseys but one spectating and one acting. During the pan, the influence of the unwanted nearer object coming into frame would be evaluated. Ideally, this requires a motorised panning head.
Years ago, I remember that consumer videocam SR tests were standardised by having the camera on a vibration table at two speeds to emulate muscle tremors and shooting a video clip of a subject from the camera. (Was it on Jim Kerr's DVSPOT website?) This wasn't very high-tech, but it was good at communicating how well the SR performed. In 2013, it shouldn't be hard to perform a standardised DSLR SR test.
Dan.