Some comments on a “breakdown” 1n Koopmans' theorem