Deligne's idea was that Shimura varieties should be understood as moduli space of motives(with extra structures). lot's of Shimura varieties of abelian type can be understood as moduli space of abelian motives and any way essentially you can study all of them in this way by using Deligne formalism about the relation between two isogenous Shimura datum.
but there are Shimura varieties that are not of abelian type. My question is if there is any such Shimura variety that has a "nice" moduli description?