The Utility of Time-Symmetric Quantum Counterfactuals: A Response to Kastner