I hate this graphic. Should we consider the 3rd dimension as well? I mean, the "thickness" of the barrel is much bigger behind e.g. UAE than Oman. From the numbers it seems we shouldn't.
But the boundary follows the arc on the top and bottom, so it's definetily projected to the half surface of a cilynder, but the voronoi lines are straight, but they should curve just like the top and bottom.
I would be legitimately impressed if the data is scaled based on the volume of the implied cylindrical section defined by that surface border, if only because that's a pretty annoying calculation to make.