Revisiting the optimal population size problem: minimal utility level and finite lives