Did you read the article? He doesn't think this would be "cool"! He's saying that this could be used to identify people who thought they were anonymous. The main point he makes is that people consider the fact identification is difficult if someone has only one location, but overlook that it's much easier if someone has two locations.
Yes I have read the article and as other people have pointed out I too am missing the threat. This is like saying I will pull out the identity of the person who browsed through my blog and left a comment. It is definitely possibly to do it over a period of repeated visits. And assuming I spend all that time to dig out information.
So, instead I am positively noting that you cannot assume it is anonymous and think of ways I would benefit from it whether identified at an aggregate level or an approximation or even uniquely.
The author listed few examples which he words as 'exploits' the non-anonymity and I mentioned few more but rather positively. It might be a very fascinating view of the web if we can start browsing it from a world map, don't you think so ?
