I had the same thought.

You'd think that knowing accurate dimensions and features of something so nearby could be used to gauge distortion on the face.

Likewise, perhaps some kind of computer vision could quickly search for the distinctive pattern of that hat in other images/video.