DXOMark and their new resolution system

With the Nikon D7000 and 16 mp tests: Using actual lw/ph from Photozone I compared the Nikon 16-85 VR with the Tamron 17-50 (new VC). DXO gives the Tamron a much higher score (9 vs 6) and yet the Nikon has better resolution figures at 16mm vs 17mm, 24mm vs 24mm, and 50mm vs 50mm. I used the figures nearest f4 and f5.6. The CA was about equal with a slight edge going to Tamron. On DXO I also got CA about equal - with a slight edge going to Nikon this time.

I am very puzzled how DXO arrived at their conclusion unless sample difference can be that great (which I doubt). The real problem with DXO (always has been) is they never tell you exactly how they get their results or give you raw data. They say they use MTF scores but these are given out by the camera companies as wide open (f stop) figures. Do they use these? Do they measure at other f stops?

And your thoughts?

--
Steve Bingham
www.dustylens.com
www.ghost-town-photography.com