Comment on "Universality of finite-size scaling: Role of the boundary conditions"