This Minireview summarizes the fundamental results of a comparative inverse-model versus real-model catalyst approach toward methanol steam reforming (MSR) on the highly CO2-selective H2-reduced states of supported Pd/ZnO, Pd/Ga2O3, and Pd/In2O3 catalysts. Our model approach was extended to the related Pd/GeO2 and Pd/SnO2 systems, which showed previously unknown MSR performance. This approach allowed us to determine salient CO2-selectivity-guiding structural and electronic effects on the molecular level, to establish a knowledge-based approach for the optimization of CO2 selectivity. Regarding the inverse-model catalysts, in situ X-ray photoelectron spectroscopy (in situ XPS) studies on near-surface intermetallic PdZn, PdGa, and PdIn phases (NSIP), as well as bulk Pd2Ga, under realistic MSR conditions were performed alongside catalytic testing. To highlight the importance of a specifically prepared bulk intermetallic[BOND]oxide interface, unsupported bulk intermetallic compounds of PdxGay were chosen as additional MSR model compounds, which allowed us to clearly deduce, for example, the water-activating role of the special Pd2Ga-β-Ga2O3 intermetallic[BOND]oxide interaction. The inverse-model studies were complemented by their related “real-model” experiments. Structure–activity and structure–selectivity correlations were performed on epitaxially ordered PdZn, Pd5Ga2, PdIn, Pd3Snv, and Pd2Ge nanoparticles that were embedded in thin crystalline films of their respective oxides. The reductively activated “thin-film model catalysts” that were prepared by sequential Pd and oxide deposition onto NaCl(001) exhibited the required large bimetal[BOND]oxide interface and the highly epitaxial ordering that was required for (HR)TEM studies and for identification of the structural and catalytic (bi)metal[BOND]support interactions. To fully understand the bimetal[BOND]support interactions in the supported systems, our studies were extended to the MeOH- and formaldehyde-reforming properties of the clean supporting oxides. From a direct comparison of the “isolated” MSR performance of the purely bimetallic surfaces to that of the “isolated” oxide surfaces and of the “bimetal[BOND]oxide contact” systems, a pronounced “bimetal[BOND]oxide synergy” toward optimum CO2 activity/selectivity was most evident. Moreover, the system-specific mechanisms that led to undesired CO formation and to spoiling of the CO2 selectivity could be extracted