Predicting the effective properties of multi-phase materials from the constituents' properties and material's morphology is a classical problem of materials science. Here, a generic finite-element based approach for predicting the behavior and properties of multi-phase materials comprised of anisotropic, arbitrarily shaped and oriented phases is presented. It is the consistent use of periodic boundary conditions in the course of generating multi-inclusion Monte Carlo configurations, dividing them into morphology-adaptive meshes, and numerical solution for the overall properties that extracts an accurate prediction of the behavior and properties of multi-phase materials from remarkably small computer models. The approach is employed to identify numerically the technological potential of some whisker and platelet filled polymers, including thin polymer films. It is demonstrated that the filled polymers studied have an appealing, currently unrealized technological potential. This potential has essentially been overlooked because of the wide-spread use of empirical equations tuned to reproduce available experimental data that unavoidably reflect the existing technological level.