How Much Could Be Gained by Optimal Stochastic Control Policies