A general notion of the simulation fidelity in system verification and validation is explained in the modeling & simulation framework of the experimental frame and discussed in the context of modeling abstractions. A formal method for fidelity measurement is presented for semantic compatibility through abstraction inclusion relations. Behavioral modeling abstractions of dynamical systems consistent with simulation objectives are explained in the experimental frame formalism using bisimulation relations and their approximations. Property-preserving bisimulation relations consistent with the simulation objectives are explained in the context of hierarchical abstractions and composition of experimental frame components. The behavioral fidelity problem is presented by finding bisimilarity preserving surjection maps of desired abstraction precision. In abstraction of linear time invariant systems, the paper discusses the problem of extending the classical results to finding surjections whose null space is a controlled invariant subspace with respect to fidelity requirements. Applicability and derivability of experimental frames are discussed through such consistent abstractions. In conclusion, challenges in research and implementation of a semi-formal unified fidelity framework encompassing both syntactic and semantic aspects are discussed. Nomenclature EF = Experimental Frame H = linear surjection map M = model X = inputs to model Y = outputs of model S = states of model = abstraction β = derivability = applicability = bisimulation function = abstraction precision = requirement = transition system