Proving the structures of charged metallacages obtained by metal ion coordination-driven solution self-assembly is challenging, and the common use of routine NMR spectroscopy and mass spectrometry is unreliable. Carefully determined diffusion coefficients from diffusion-ordered proton magnetic resonance (DOSY NMR) for six cages of widely differing sizes lead us to propose a structural reassignment of two molecular cages from a previously favored trimer to a pentamer or hexamer, and another from a trimer to a much higher oligomer, possibly an intriguing tetradecamer. In the former case, strong support for the reassignment to a larger cage is provided by an observation of a slow reversible transformation of the initially formed cage into a smaller but spectrally very similar one upon dilution. In the latter case, freeze-fracture transmission electron micrographs demonstrate that at least some of the solutions are colloidal, and high-resolution electron transmission and atomic force microscopy images are compatible with a tetradecamer but not a trimer. Comparison of solute partial molar volumes deduced from measurement of solution density with volumes anticipated from molecular models argues strongly against the presence of large voids (solvent vapor bubbles) in cages dissolved in nitromethane. The presence of bubbles was previously proposed in an attempt to account for the bilinear nature of the Eyring plot of the rate constant for pyridine ligand edge exchange reaction in one of the cages and for the unusual activation parameters in the high temperature regime. An alternative interpretation is proposed now.