Optimal Control Problems in a Nonreflexive Framework and Application to Weakly Converging Nonlinear Input Operators