A Critique of Rees's Theory of Primordial Gravitational Radiation