Thanks guys, and I'd like to add a follow-up comment here. To be clear, I still see improvement in detail going from the 5D to the 1Ds2 to the 1Ds3, however, to my eye it is even less significant at ISO 1600 than it was at ISO 400. But it is there. As to whether the difference/improvement is significant enough for the cost deltas, folks need to evaluate that for their own uses.
When the 5D was released a few years ago, I said pretty much the same thing about the 1Ds2, but back then felt the 1Ds2 was worth it for my uses; I could see the difference in fine detail in my large prints and at that time in my photographic career, optimum detail was the holy grail for me. However, it wasn't long before I noticed that if I wasn't comparing the 5D print and the 1Ds2 print side-by-side, both prints looked good to me. Sure, I could see improved fine detail in the 1Ds2 prints, but when I asked several non-photographers in my building which print they liked better, I got varied answers. (My usual print buyer is a non-photographer.) The common comment was, "They look the same to me," or "I like the colors in this one better." And it was a toss-up as to which print they pointed to when they made the color comment...
That was when I realized that image content, not detail, is what made my photographs "purchase worthy," and I started focusing my energies trying to make better images, not just more detailed ones. (Edit: I would like to point to MR's crop of "Mango Langway" in his Madagascar series: Nearly void of detail in the technical digital camera sense, yet it very definitely works as an artistic image.) And thus is how my own personal approach to gear needs has evolved over the past few years. But, just because that works for me, doesn't mean it will work for everybody, and I post my camera comparisons in that spirit --- hopefully I am presenting the data in a fashion that allows everybody to make a more informed decision for themselves.
Cheers,