We previously reported on a thermal regime where pure, polycrystalline methane hydrate is preserved metastably in bulk at up to 75 K above its nominal temperature stability limit of 193 K at 0.1 MPa, following rapid release of the sample pore pressure. Large fractions (>50 vol.%) of methane hydrate can be preserved for 2-3 weeks by this method, reflecting the greatly suppressed rates of dissociation that characterize this "anomalous preservation" regime. This behavior contrasts that exhibited by methane hydrate at both colder (193-240 K) and warmer (272-290 K) isothermal test conditions, where dissociation rates increase monotonically with increasing temperature. Here, we report on recent experiments that further investigate the effects of temperature, pressure, and composition on anomalous preservation behavior. All tests conducted on sI methane hydrate yielded self-consistent results that confirm the highly temperature-sensitive but reproducible nature of anomalous preservation behavior. Temperature-stepping experiments conducted between 250 and 268 K corroborate the relative rates measured previously in isothermal preservation tests, and elevated pore-pressure tests showed that, as expected, dissociation rates are further reduced with increasing pressure. Surprisingly, sII methane-ethane hydrate was found to exhibit no comparable preservation effect when rapidly depressurized at 268 K, even though it is thermodynamically stable at higher temperatures and lower pressures than sI methane hydrate. These results, coupled with SEM imaging of quenched sample material from a variety of dissociation tests, strongly support our earlier arguments that ice-"shielding" effects provided by partial dissociation along hydrate grain surfaces do not serve as the primary mechanism for anomalous preservation. The underlying physical-chemistry mechanism(s) of anomalous preservation remains elusive, but appears to be based more on textural or morphological changes within the hydrate material itself, rather than on compositional zoning or ice-rind development.