Discussion of “High-dimensional autocovariance matrices and optimal linear prediction”