CCL:G: Translational entropy in solvent

Dear CCLers:
The problem of the entropy component in solvent was tackled for instance in J. Am. Chem. Soc. 2004, 126, 10457 and Organometallics 2006, 25, 3647 by simply deleting the rotational and translational entropy. This extreme approach can be sometimes more accurate than putting all the free energy components in the same basket with the solvation. Comparing my simple gaussian values with the more accurate ones of J. Chem Phys 107, 1997, 1981, convinced me that for charged molecules in high dielectric solvents it's better to delete the entropy in gas phase.
For a Ni neutral complex in THF solvent I found that the accurate values where roughly in the middle between full free energy and E+ZPE. This compared to experimental values and considering my DFT calculations accurate enough. Clearly gaussian overestimates the translational and rotational entropies. It would have been nice a manual that explains this behaviour. I've heard of programs that correct this, but still I couldn't find a good manual for them also.

..........Sebastian Kozuch...........
...The Lise Meitner-Minerva Center...
.for Computational Quantum Chemistry.
...Hebrew University of Jerusalem....