In spite of free-atom electronic-relaxation contributions to transition-metal cohesive-energies, numerous studies have misused the latter instead of using the solid-state interatomic bondenergy in modeling bulk and surface properties. This work reveals that eliminating the free-atom contributions from experimental cohesive-energies leads to highly accurate linear correlations of the resultant bond-energies with melting temperatures and enthalpies, as well as with inverse thermal-expansion coefficients, specifically for the fcc transition-metals. Likewise, predictions of surface segregation phenomena in Cu-Pd and Au-Pd alloys on the basis of the modified energetics are in much better agreement with reported low-energy ion scattering spectroscopy (LEISS) experimental results, as compared to the use of cohesiveenergy values. A last demonstration of the problem and its solution involves the significant impact of the modification on segregation (separation) phase transitions in Cu-Ni model nanoparticles.