The screening for HP is done at 8 bit for the Z and 16 bit for the 9180. Bit depth for screening is not the same as image bit depth. Epson it's done in the printer, I have no idea on their actual bit depth of masking. I think it is hard to argue that the finesse of the Epson screening is extremely fine, and efficient. In fact it is too real for many of us used to film, press dots etc.
I didn't ask what the processor ASIC has in bit depth but if the new 11880 16 bit drivers produce visual differences.
The reasons of the mild clipping in Epson drivers is not at all bit depth of masking but a decision of pleasing contrast and sharpness. Easy enough to see if you use GutenPrint to run a raw plot.
You are combining a lot of things that are related to total output to visual response to Granger rainbows. I use the most tools as indicators and thus don't have as much conviction as your might due to the variability of the tests I worked on and will work on in the future.
Do you have some documentation on Canon using 12 or 10 bit for masking. I haven't seen any yet, but an awful lot of assumptions either way. When I asked Canon engineers from Japan, they just wouldn't say. Others such as John Panazzo asked too, same they didn't want to say. At least HP is clear and told me quite a bit about the masking other things that are confidential; of course.
Although one would quickly assume that higher bit processing to be an advantage, fact is it is required for performance within the hardware set up to perform at it's optimum , which in the case of Canon is a good thing. Yet assuming the same for other systems is not true until proven , and that, is something that would be for R&D to do.
I don't know what EOM profiles are. OEM? PMP is decidedly different in skin tone repro. I think Monaco is too yellow for my taste, too yellow for a lot of printers too. In the fashion /beauty world and or Asiatic side a brighter magenta is definitely the side you want to be on. The number of times press runs are credited back to the big names have most often been skin tones being too yellow, too dark etc. Considering the Logo guys and girls are over here, I think the skin tones they look to are pointed towards the look pros are looking for. If the labs there are having enormous problems with skin tones I question their source images more than PMP.
We're getting too far off topic though , as all of us have preferences as well as differences in perception/vision. I have always loved discussing these subtleties! Just a note on large groups> there was one image in a lab test with a girl in a bathing suit (hmmm my type of image for testing/ eval !) that unanimously ( a lot of pro photogs) was better on the Epson than Canon or HP. The skin tones were yellower, redder and more saturated, the bathing suit much richer in red and darker thus higher local contrast. Thing is , the perfectly calibrated screen, L*a*b* eyedropper readouts, and Canon and HP ( HP was better) prints were the correct match.