the differences are not induced by the colorimeters itself (just partial) but by the way the calibration softwares work up (interpret) the measured data.
Thats a good point as well. I've noticed that myself as Eye One Match and Color Eyes Display Pro have different versions of the same white point with the same puck. The difference isn't as big as the difference from different manufacturer's puck to different manufacturer's puck but it's a colorimetrically significant difference such that if you created profiles to a given color temperature using the same puck and two different software packages, you'd have two very different profiles (of course, even if you measure the same white point as a reference before profiling you'll still have two very different profiles but that's a different discussion).
It's a damn shame that something as important as a properly managed color workflow isn't as glamorous as megapixels and the latest software update that ultimately doesn't do a whole lot more for your workflow than the previous version.
I typically work with a group of people (wedding photographers... I am one myself) who will spend tens of thousands on camera bodies, lenses and fast computers but when it comes to their monitor and calibration system they really don't want to spend more than $500 bucks on the ONLY thing (before they get an album that costs $550 to produce back from the lab) that is going to tell them what their work is going to look like when it hits a lab. Heck, when I say "soft proofing" most of them give me a blank stare.
I wish I could figure out how to make something that would cost a full time wedding photographer $23.30 per wedding over the course of 3 years (and will ultimately save them a lot of time and money) more glamorous instead of buying the latest whiz bang action set. Okay... now I'm off on a tangent.