Adjust the final futility bound with same value as the efficacy bound for asymmetric non-binding design

[LittleBeannie]

When we use integer design with gsDesign2 and asymmetric non-binding bounds the final efficacy and futility bounds are not the same. In gsDesign they are the same. In gsDesign, we make the final bounds equal, the efficacy and futility bounds unequal and get Type II error for the futility to match specified value better. My usual practice is just to eliminate final futility bound.