Absorption chillers, heat pumps and heat transformers exploit thermal power input to satisfy a variety of refrigeration and heating applications, as well as to boost the temperature of low-grade heat. Can one arrive at accurate predictions of system performance with a simple analytic irreversible thermodynamic model? In capturing the essential physics of the problem, that model would be required to provide a predictive and diagnostic tool and to permit determination of optimal absorption system operating conditions. We derive such a model and check its validity against experimental data and computer simulation results from a variety of commercial absorption units. We reinforce the observation that commercial units appear to have evolved empirically such that maximum efficiency is realized under design conditions. The failure of a host of previously-published endo-reversible thermodynamic models to account for fundamental qualitative features and accurate predictions of system behavior is documented with computer simulation results and experimental performance data.