I just tried your link. Here are my results: "Your browser fingerprint appears to be unique among the 2,399,190 tested so far."
I'm one in roughly 2.4 million. That's untraceable enough for me :p
1 in 2.4m is being similar to ~150 other Americans. Not good odds for hiding among the crowd once non-client browser specific identifying clues are taken into account, such as browsing habits. 1 in 2.4m is being similar to ~400 other facebook users, and facebook knows a lot more about its users than just the client browser fingerprint.
That isn't 1:2.4e6, that is at best 1:2.4e6, they have only run 2.4 million tests and so can't say anything more than that. Your browser fingerprint could easily could be unique in the whole world.
Until I change just one of those identifiable metrics, that is, right? As soon as I change, for instance, my screen resolution I am another unique user and am no longer identified with the previously unique fingerprint.
Surely, if someone installed insidious software on every single page of the internet and tracked me everywhere, there could be a way to use some fuzzy logic to re-identify myself. But we're way into the deep end on hypothetical again, right?
My main point is that, practically speaking, it isn't very.