I have exactly the same problem. I tried with several LCD screens
(CFL, LED) and the readings were still too blue which resulted in too
red ICC (compared to Spyder 2 results). I created CCMX for my colorhug
(Lenovo T500 and i1 Pro spectrophotometer was used for this task):
http://jskarvad.fedorapeople.org/lenovo_t500_colorhug.ccmx. It helps a
lot, but the results still seem not to be optimal - if I lower the LCD
backlight the readings are still too blue even with this CCMX. I don't
know what's going there, but I will try to experiment more.