Hi Jack, I have the camera and have shot it side by side outdoors where I believe I can make a pretty good educated guess. Hands down the Pixel shift shows considerably more resolution throughout a frame, compared to a D810, same sensor.
But I did review your detailed thread this morning.
You make a good point on the extra time taken by the camera, and needing to be on a tripod as I did not mention that. Camera takes about 4 to 5 seconds totally to finish the job. File size is not a big issue for me, as I am still shooting Phase One also and used to the large files.
What started me that direction was the simple test that Dpreview did months ago, then the Imaging Resource tests (D810 5dsr and K1). The Dpreview static tests have always been very good at least for me, and the comparison of the K1 in Pixel shift mode at higher ISO's was most impressive. However now that I have the camera and have made quite a few trips with it, the Pixel shift images are so much cleaner even at base ISO, (shadows and haze etc) it's hard not to want to shoot this way each time. It's like a big weight just was lifted off my shoulders.
My only issue with the K1, is the way it handles highlights, and that is very tricky. Nothing like the Nikon D810 at base ISO, so I have been burned a few times no pun intended.
The camera is about the same in noise without pixel shift as the D810, the nod going to the D810 at ISO 64. But when you move to pixel shift, all bets are off for the Nikon, its just that much better.
Also where the differences really start to show, is take two images non pixel and pixel shift then uprez them with any of the software tools out there to around a 30 x 44 at 300 dpi. The pixel shift image handles the uprez much better at least for me. For my testing I used Photozoom Pro and LR.
Paul C