Scott Garrabrant comments on Biextensional Equivalence