Comment on ``Testing integrability with a single bit of quantum information''

In quant-ph/0303042, Poulin, Laflamme, Milburn and Paz consider the problem of distinguishing quantum chaos from quantum integrability for dynamics in an $N$-dimensional Hilbert space. They claim that this can be done by deterministic quantum computing with a single bit using $O(\sqrt{N})$ physical resources, compared to O(N) physical resources classically. I point out what seems to be a fatal flaw with their proposal.