The liquid drop (LD) model is revisited to assess the reliability of its predictions for thermodynamic properties of cluster ions and to examine the rate of convergence of such properties to their bulk counterparts. The model predictions are in very good agreement with both available experimental data and simulation results for Na+(H2O)(n) clusters, surprisingly for all cluster sizes, and the stepwise cluster thermodynamic properties are found to converge only slowly to their bulk counterparts. The LD model allows a natural partitioning of the cluster ion thermodynamic properties into various components, one of which (the solvation part) is of prime importance in connecting cluster solvation properties to the bulk limit. The latter LD model component is found to be entirely analogous to the so-called dielectric sphere theory, and as implied in earlier work, the results of dielectric models suggest that ion solvation is also a very slow process to converge to the bulk limit. In addition, a form alternative to the customary interior ion LD model is proposed, where the ion resides at the surface of a solvent droplet, and the resulting model successfully predicts that surface ion I-(H2O), clusters are thermodynamically very likely and that large halide ions tend to be located, if not at the surface, very close to it in large clusters of polar solvent molecules. Conversely, small ions such as Na+ are predicted to be interior in water clusters. Further, large ions such as I- are predicted to have interior but near-surface locations in acetonitrile clusters. Even though it seems to work better for smaller ions and solvents such as water, the LD model, despite its simplicity, generally appears to properly describe cluster ion thermodynamic properties over a wide range of cluster sizes and even for relatively small cluster sizes.