.....IMO it is dangerous to generalise about the reliability of models, but it seems to me that the key issue to focus on is whether models rely extensively on people switching modes of transport or whether the traffic is mainly expected to come from merely diverting journeys that use the same mode.