Ray, hate to say it, but though you seem to otherwise be a reasonable sort of fellow, on this subject you're simply wrong. This comment is based on my intensive experience comparing the D3 and the 5D and 1Ds MKIII, and that of quite a few other photographers who have done detailed side-by-side comparisons of the D3 at high ISO against Canon's best.
You may decide that we are all wrong (and this includes probably half a dozen reputable magazine and web reviewers as well, many of whom I know to be level headed and knowledgeable people) but then again maybe, just maybe, it's you that isn't seeing things for what they are.
You can push process Canon files till the cows come home, and you can claim that the D3 smoothes detail, but neither of these changes the fact that knowledgeable observers seem to be in agreement that there is at least a two stop advantage to the D3, and that it's real, quantifiable, verifiable and there for anyone with an experienced eye to see for themselves.
But, you are of course free to believe whatever you wish. It seems pointless though to continue to argue otherwise here since all that we have from you is your opinion on the matter, while the other side of the argument has experienced observers on three continents claiming otherwise.
Ps: I would add that "cleaning up a 5D image in Noise Ninja" to make it competitive is a red herring, simply because one can then do the same thing to a D3 and still have the two stop advantage.
As for quoting ken Rockwell on the matter, well, what can I say? I should simply leave it at that.