I would also wonder why not just printing from Photoshop. I don't myself, but that is because I print from Qimage. What I do, however, is soft proof before going to Qimage to see which way will be better, either relative colorimetric, or perceptual. Depending on the image, one may be quite better/worse than the other.
I'm also somewhat curious about step 9. If you have a well calibrated system, I don't see the need for this. Aren't you starting to "close your eyes" here? If you get results you like, well, great. But to me, the image should be able to be seen on screen accurately to how it will print. Is there some other dynamic at work that I'm not understanding?