Digital BW, The Print

This past weekend I tried the curves and workflow created by Paul 
Roark found on the MIS website for the UT-FSN inkset on an 1160.

I found the result quite terrible and the instructions confusing, so 
apparently I'm not doing something correctly.

I was using EEM setting my color space to Adobe 1998 and found that 1 
curve gave extremely posterized results, and two other curves 
lightened the print way too much and dark tones looked as though 
there was a veil over those areas.

By comparison when using the R9/Cone driver and the John Wolfe work 
flow results are much better, Wolfe's workflow produced the best 
print by far, WYSIWYG in fact and I'm using a calibrated and profiled 
monitor.

Have other people using the Roark curves, etc. with UT-FSN 
experienced the same results?

Thanks,

Don Bryant