Debiased Inference of Average Partial Effects in Single-Index Models: Comment on Wooldridge and Zhu