For the life of me I cant figure out why so many of you are saying there will not be an IQ difference on a coffee table book if shot on a 35mm 23mp camera vs a larger 60mp sensor....
Must get to the bottom of this!
Join the circle of confusion
It's really about the output. The best quality commercial offset printing is done via halftone screening at ~200 lines per inch (LPI) - that's for art books, as opposed to magazines and newsprint which use lower resolution screens (in the 80-150 LPI range). Printers generally want twice that in dots per inch (DPI) of digital resolution.
Most coffee table books that I have and have seen are square. That means a 2-page spread will have an aspect ratio between 2:1 and 2.4:1 (allowing a 1" border with no gutter, bigger books get closer to 2:1). To cover that aspect ratio, you'd need a 4-frame stitch from your 3:2 sensor in portrait orientation (~3.3 frames needed, but you always need overlap to stitch). With the 22 MP of the 5DIII, that gives you ~70 MP final image. A 2-page spread of a 12x12" coffee table book means a 22x10" image, so at 400 dpi you need 8800x4000 pixels, or 35 MP…you'd have twice the amount of resolution you need. That 70 MP image would be sufficient for a 2-page spread in a coffee table book of up to about 16x16".
If you need to go larger, you can use a longer focal length and stitch a two-row pano together, although that gets a little more complicated (you need a multirow pano setup, rather than just a single-row). You can put a 1.x4III behind your TS-E 24L II, and still have excellent resolution and IQ from the combo.
Neuro...as always, a clear and succinct explanation of the problem. Thank you for posting. We all learn so much from you on this site.