Motion perception deficit: risk factor or non-specific marker for neuro-developmental disorders?

In their comprehensive review of motion perception in autistic spectrum disorder (ASD), Milne and colleagues (2005) mention both that deficits in motion perception are found in only a subset of autistic individuals, and that such deficits have also been evidenced in other neurodevelopmental disorders than ASD. I think that this point deserves further discussion. Indeed, it bears quite directly on the issue whether a motion perception deficit reflects an underlying cause of ASD or not. If it does, how can one explain that it is not present in all autistic children, and that it is present in children with other disorders but no autistic symptom? If it does not, why is it associated with ASD (and other developmental disorders)? Explaining this paradox requires a relatively sophisticated causal model. I will describe and discuss the respective merits of two classes of models that may serve this purpose: the multiple risk factor model and the non-specific marker model.