On the origin of the extremely different solubilities of polyethers in water