Eyeballing the lower and upper quartile looks like the shape of the front page distribution is roughly the same. Now if we do this for subsequent pages and can fit past and present shapes with linear scaling, I'd say there's nothing to concerned about here.