We apply local orbital basis density functional theory (using SIESTA) coupled with a mapping to the Anderson impurity model to estimate the Coulomb assisted or correlated hybridization between transition-metal orbitals and ligand orbitals for a number of molecular complexes. We find remarkably high values which can have several physical implications including (i) renormalization of effective single-band or multiband Hubbard model parameters for the cuprates and, potentially, elemental iron, and (ii) spin polarizing molecular transistors.