>You also did not mention the distance you viewed the prints in >the various lighting, and any optical aids you used?
Well Neil, while I did mention my viewing distance, I did not specifically reference it when discussing my print evaluation. I was using my prescription cheaters at the closest viewing distance I could comfortable view the print, which was about 15 inches which is the average viewing distance for that size print. At that distance I believe the average person can discern about 458 PPI. In my further testing I then used a magnifying glass and even under the incandescent, I could then determine the 300 PPI print as having more detail. But that is akin to putting your nose to the print without any aid.
I also agree that there is no one size fits all answer here. For example many dye-sub printers and machines that produce tradition wet processed prints top out at 285 to 314 PPI. So any image resolution above that is wasted. There are higher end units with greater PPI ability (600 PPI for dye-subs, 4000 PPI for image setters). In this case I used PPI instead of DPI for those processes are continuous tone (three printer dots overlaid which produces all the color information for a pixel).
For inkjets it gets quite complicated because of their fixed ink color set and stochastic pattern they use. They have a high DPI if printing monotone, but when printing con-tone they need multiple dots to simulate one pixel. It’s also compounded by the fact that pixels are square, printer dots are round. At higher resolutions the space between dots allows for a smoother transition and the pixels become co-merged and no longer recognizable as square. But at lower resolution enough dots can be used that the pixel becomes obviously square, which then takes on that stair-stepped look to specific detail. It was my determination that this occurred at about 100 to 120 PPI with my HP printer.
Then there can be a disparity between hardware and software resolution. For example Rags Gardner in his article states that when studying the documentation for the Fuji Lightjet, he discovered the laser could achieve 317.5 PPI, but the RIP was working at 300 PPI. In that case there was no optimum PPI to use due to the disparity between hardware and software.
In the end, I agree with your insistence that everyone research and test this out for themselves, for there are just too many variables involved to give a generalized answer.