There's been very little work done on Four-posters, so we don't know much about them. Either a favourite way of knacking them was by breaking the symmetry and removing one of the boulders or else there were many 'Three-posters' built, which doesn't make much sense. The best preserved example I know has two pairs of low outliers, one upstream and one downstream, and it seems reasonable to assume that they are 'under-represented in the historical record'.