Approximated closed-form minimum-cost solution to the (r, q) policy with complete backordering and further developments