1: Most papers will respond to the same ink limits differently so the generic cross over points in the raw profiles might not (more likely wont) be accurate from one paper to the next. When using only three gray inks it might not be absolutely critical to have the exact cross over point from one paper to the next, but it can make linearizing easier in keeping down the "not constantly increasing" errors down.
2: Some papers might do better with a higher or lower K limit and K boost. You will see that in the measurement of the 21 step file it if stops increasing past the 90% patch. If it doesn't stop then how do you know how dense you really can make it? I've pretty much settled on keeping the ink limits for K, Lk, and LLk between 55 and 64 and then adjusting the boost depending on the paper. Changing those K limits would require changing the cross over points for the LK and LLK inks too (gray_value= in the text file or the "density" inputs on the windows gui). To do that you would need still need to print the ink separation page for the K3 inks with those new ink limits, and then follow the steps for interpolating the cross over points.
The bottom line with not printing and measuring the ink sep page is you don't know what ink load your paper is capable of holding.
Richard Boutwell