Digital BW, The Print

I've just started playing around with the UT-3D inkset in my Epson 1290 
and have made a few PS curves for EEM. In preparation for some QTR 
experiments I printed the QTR calibration page, and to my great 
surprise the primary toner (magenta and light magenta position) inks 
look quite different to each other. Whereas the LM position ink is 
distinctly bluish, the M position is much more of a cool gray - it 
certainly doesn't look like a continuation of the same shade through 
the darker tones, the way the carbon inks do.

I didn't expect this. Has anyone else had the same experience? Is my 
assumption correct that the colors in the full-strength and light 
positions would look roughly the same? Or did I just not shake the 
bottle?  ;-)

Thanks in advance,
Ben Connor

P.S. If anyone else is using, or plans to use, this inkset with this 
printer, I'm very happy to share my curves or to try yours. Drop me a 
note if you're interested.
BC