Also (with the hope of not starting a war!) the following remarks on lens testing may be of interest:
I agree with this view expressed by Bob Atkins. Test methodology does vary. Lens quality from batch to batch does vary. Subjective opinions vary most of all.
The solution to these problems appears to be too expensive and I'm not sure that manufacturers want solutions. Manufacturers like brand loyalty. I don't believe they would welcome a method which would make it easy for critical and objective scrutiny of the performance of their lenses by consumers.
I believe there's a statement on Photozone's site to the effect that they reserve the right to remove any test report which they subsequently believe is a-typical of the performance of that lens.
In my own situation, I feel I would like at least one other lens to go with my D700. I'm very pleased with my Canon 50/1.8 because it's very lightweight and remarkably sharp for its price. I imagined the Nikkor 50/1.8 would be similarly good value. However, when I looked at the Photozone results, I almost fell off my chair. Fortunately, I didn't bang my head.
The Nikkor 50/1.8 certainly seems sharp in the centre, perhaps even sharper than the Canon 50/1.8, although one has to bear in mind that the Canon test was performed with an 8mp camera (the 350D) and the Nikon test with the 10mp D200. I think it would be fair and reasonable to add 10% to the Canon figures for comparison purposes.
But what caused me to almost fall off my chair was edge and corner resolution. Is this a lens designed for full frame? If it is, how can it be that bad at the borders? If the borders are that
bad on a cropped format camera such as the D200, what are they going to be like on a full frame like the D700? Well, one could reasonably assume, much worse.
You judge for yourself. I post the Photozone resolution charts for both lenses below. Hope Photozone doesn't mind.