This would be interesting if it aggregated a lot of sites across segments. Maybe there is only one Music site in the top 100, but the aggregate for Music may w

This would be interesting if it aggregated a lot of sites across segments. Maybe there is only one Music site in the top 100, but the aggregate for Music may well be much higher. That is, what's the aggregate for each segment, what's the average and the stddev. From this you'd see what segments are more fragmented than others -- i.e. there might only be two "Auction" sites because eBay has it sewn up. There might be only one music site, because there are 1,000s of music sites and nobody's dominated.