I did indeed misunderstand the statement of Friedman's Theorem 2, which asserts (if I have it right this time) that NBGC + measurable is interpretable in NBG + powershift, not vice versa as I read it! --Randall Holmes