I don't mean to be cranky, but why do some people fail to comprehend that all that was tested here was resolving power? Wasn't that clearly enough stated in the article?
The tonal smoothness and absense of grain at low to moderate ISO is a huge factor to consider when comparing image quality between systems. It is the case now and it was the case 6 years ago with the D30.
It was only 3MP, but its images were so clean that as long prints were within the constraints of its file size, film had met its match. I said it then and I say it now. But people love to take that statement out of context, and it's grown really tiresome.
(If this hadn't been the case we'd all still be shooting Velvia, now wouldn't we?)
Image quality is a composite of a great many factors – file size, image size, resolution, accurance, colour accuracy and rendition, noise floor, highlight retention, dynamic range, variable ISO noise characteristics, long exposure noise characteristics, and on and on.
Perspective folks. Keep perspective.