The Argyll site still seems to be down but I am interested, you don't seem to convinced, what didn't you like?
Is making a profile not purely mathematical what would make one software better at measuring a target and calculating the differences than another?
I have looked at ColorMunki and Spyder3Print and cant decide, but at the moment I just need to do some experimenting with a scanner profile mostly for reflective scanning.
Making a profile is mathematically easy, I believe, if you have lots of patches in the target; however that is not a situation which is very pleasant when you have to handread the patches. The nice thing about ColorMunki is that it makes good profiles from just a few patches.
Regarding Argyll, I wrote a Linux wrapper demo for Barbieri for their LFP instrument, and the RGB profiles I generate with 1400 patches and Argyll are competitive *to my eyes* with the profiles I generated using an Xrite iSis XL, Xrite's Profile Maker Pro software and the 4000 patch Atkinson targets. In any case they are what I use for my own gallery work. But I did have to invest a few weeks in getting things to work like I need them. I'm not going to say anything bad about Argyll, it is maturing into a nice profiling tool and it can do a lot of things which no other tool can do.
However I would still recommend that anyone needing to profile for quick inkjet art printing just get a ColorMunki at this point, or get a consultant (eg. Andrew, Ethan or me) to generate profiles.
Edmund