The problem with this approach is that there are some company and school grids that are very large, without being social virtual worlds.
Take the MOSES grid, for example, which has around 200 regions, and a lot of users -- miliitary, researchers, academics.
To me, this is a prime example of the kind of uses OpenSim is going to be seeing a lot of. In order to get into the grid, you have to be approved for access, to keep out griefers -- and maybe foreign spies. But almost anyone else can get permission to visit, to see what they're doing.
But many commercial social grids also have a procedure where they have to approve membership applications.
Right now, commercial social worlds make up a small fraction of all the grids -- just as the social websites like Facebook make up a small fraction of all the websites.
To me, a "grid" is any immersive virtual environment laid out in a checkerboard pattern (as opposed to room-based virtual environments with no map, only teleport connections).
A "commercial social grid" or "commercial social virtual world" is one that includes all the functionality you're looking for -- basically, Second Life clones.
I understand that a lot of people are looking for that, and such a list is useful.
But it's not my primary area of interest.
I do write about such grids when they have new business models -- I wrote about the sex grid because it was a new business model. I wrote about AviWorld last year when it launched with the idea that it would charge more for land than everyone else (though that business model didn't work out). I write about Kitely a lot because their business model is unique and, if it works out, I expect a lot of other grids to follow suit.
But I don't write an article every single time a new social grid launches. Usually, I just mention them in my monthly report.
I do care, very very much, about all the small, school and enterprise deployments. I think that is where the vast majority of OpenSim use is going to be. And that's why I track those stats.