I am giving an alternate proof of your implicit statement "There exists a line which has the same number of points on each side". You did this by computing duals of cell arrangements. I am arguing that you don't need to do that.

The proof I outlined will work for any point. Initially it might have the wrong number of points on both sides but for some rotation it will have the correct number of points on both sides.