Hi,
This is going to be a tricky question to phrase given that I am not too familiar with the subject matter. I was discussing Gaussian results with one of the earlier developers last week. We (more from his side) got on the discussion of the error in reported energy being scrambled/nullified due to "soft harmonic oscillators" in the vibration of some groups, in particular rotating groups.
Could anyone with a lot more experience than I, expand on what I think he was discussing, and perhaps point me towards some literature.