How close can pure functional programming get to C?

Richard Feldman digs into performance of pure FP versus C, and where the gaps actually come from (allocation, in-place mutation, etc). Worth reading if you've ever said 'FP is slow' at a meetup.