Discussion: Inference and Interpretability Considerations in Frequentist Model Averaging and Selection