If you can live with a 1:3 crop factor on wides, it's no doubt the ideal solution. Especially if Live view works in low light without excessive noise artifacts like the Nikon implementation on the D800.
In my workflow I am most often at 28mm and or 40mm and need all of it. Even in stitching. The crop factor is too limiting for me.
The Rod 32 may be a great lens on this back, but it's cost is around 9.5K new and 7K to 6K used. Based on the fact that the Copal mount is very delicate and the lens is very heavy, it's not the best lens IMO to carry around in the woods. If you are on the 1:3 sensor I am assuming that a lens with a 70mm image circle may not do much for you, and you need to be at 90mm at least, which bring up either the 32mm or 40mm.
I have long considered the 32mm, but the price it just too extreme for me. If you are considering a tech camera which again is about a 6 to 7K up front investment, then add the cost of the 32mm Rod, you are talking a lot of cash up front. Where as the deals on a used IQ160 may be a much better overall solution as you can still use the Schneider wides as much as 8mm of horizontal movement.
As far as limited 4mm movements, I can't see a price justification for that, but that's just me. I want 15mm to 25mm of movement for my work. Rise and fall are not as important, but I do use them at times.
It's still a huge deal either way for Phase One and I am sure the start of things to come. Right now, there is not a Nikon lens I know of that can stand up to 10mm of shift (for sure the Nikkor 24 TS-E can't on the D800) at least the ones I tried did not. Sure Canon has some excellent TS-E lenses in the 17mm and 24mm, but they don't have the MP. Still sitting at 21MP.
What will be most interesting will be the CI or others who test the Alpa FPS on the Canon TSE lenses. Here I can see this solution really being a great winner.
Biggest issue as far as the acceptance is that general consensus is CMOS=CHEAP, Sony led this charge also, with the Nikon D800 Chip and then newer version of same chip in the A7r. However the fab process on this chip is a bit different and I would be interested in knowing what the ratio is of good vs bad chips. Sony so far has two players for this new 50MP chip, one is already shipping, the other will be in March. Both players combined shipments for the 2014 year will more than likely will not equal the sale for Sony on one month of 36MP chips.
Still need to see how the CMOS handles extreme shifts as the one thing the LCC won't show is increase in noise vs loss of color/sat and detail smearing.
Paul