On the assumptions underlying the near-resonant theory of energy transfer