Pathfinder_77
While I can see some point in rankings, I realy don't understand b-school clusterings. Certainly, there are some criteria which distinguish Kellogg from Tuck, that's the reason why there is couple of positions in rankings between those two schools. But saying that Tuck and Kellogg are in different leagues is somehow immature and more like "BW style" posting.
What is my point? Well, there are greater differences between the schools within the same cluster than between the schools belonging to different clusters. For example, majority of the the members of this forum will agree that Stanford is >>> MIT, and they are in the same, ultra elite group. On the other hand, what is the difference between, say, MIT and Haas? MIT > Haas? Or MIT = Haas? Probably there is plenty valid arguments that could justufy one or another, but it is obvious that there is no such a large gap between MIT and Haas like between Stanford GSB and MIT. Certain applicant, lucky enough to get accepted both to Haas and MIT, would have to carefully determine pros and cons of each school and than make the choice based, most probably, on "fit" criteria and personal reasons.
I think that the point of the clustering is just as you stated: to show that there are no large gaps between schools, whereas rankings perpetuate the idea that being ranked 5th vs. 8th means the 5th ranked school is the better school, when in reality what made it ranked higher may be irrelevant to you. I think the concept is better crystallized by a comparison often repeated in the elite category. What school is better: Ross vs. Darden? Can one really argue one way or another? Or can it be said that they are (relatively) equal and that fit/personal reasons will be the deciding factor.
But you are right that at what are the perceived edges of the cluster, there will be a blurring into the next cluster. I think that is something that, if recognized properly, is a flaw worth living with.