Selective absorbers and angular selective windows use cermets and many of these surfaces have columnar features, for example, Ni pigmented aluminium oxide, many types of black chrome and silver/aluminium oxide angular selective films. There are numerous examples where optical properties in these films have been modelled with quite simple effective medium theories. Extension to optimise film designs prior to production is then assumed possible. Recent work on effective medium models has indicated why the simple models often appear to give good fits to cermet data for solar materials. Visible and near infra-red data with fitting parameters, fill factor and depolarisation factors of the variables, may need to be reevaluated. More rigorous theories show a good fit with an equation of the same mathematical form as the Maxwell-Garnett model is often expected and that parameters used to obtain these fits are not those of the simple models. Relatively simple mathematical equations for modelling is useful but compositions and particle shapes noticeably different from those assuming the conventional dipole models are possible. We use a normal polarisation modal analysis which allows us to introduce by analogy with the conventional models an "apparent fill factor". This is less than the real fill factor as a result of interparticle interactions which effectively diminish the impact of individual particles.