We are getting different results from Java for polycyclic species such as the following:
http://dev.rmg.mit.edu/database/thermo/molecule/1%20%20C%200%20%7B2,S%7D%20%7B4,S%7D%20%7B8,S%7D%20%7B11,S%7D%0A2%20%20C%200%20%7B1,S%7D%20%7B5,S%7D%20%7B6,S%7D%20%7B12,S%7D%0A3%20%20C%200%20%7B4,S%7D%20%7B5,S%7D%20%7B7,S%7D%20%7B13,S%7D%0A4%20%20C%200%20%7B1,S%7D%20%7B3,S%7D%20%7B9,S%7D%20%7B14,S%7D%0A5%20%20C%200%20%7B2,S%7D%20%7B3,S%7D%20%7B15,S%7D%20%7B16,S%7D%0A6%20%20C%200%20%7B2,S%7D%20%7B7,S%7D%20%7B17,S%7D%20%7B18,S%7D%0A7%20%20C%200%20%7B3,S%7D%20%7B6,S%7D%20%7B19,S%7D%20%7B20,S%7D%0A8%20%20C%200%20%7B1,S%7D%20%7B10,S%7D%20%7B21,S%7D%20%7B22,S%7D%0A9%20%20C%200%20%7B4,S%7D%20%7B10,D%7D%20%7B23,S%7D%0A10%20C%200%20%7B8,S%7D%20%7B9,D%7D%20%7B24,S%7D%0A11%20H%200%20%7B1,S%7D%0A12%20H%200%20%7B2,S%7D%0A13%20H%200%20%7B3,S%7D%0A14%20H%200%20%7B4,S%7D%0A15%20H%200%20%7B5,S%7D%0A16%20H%200%20%7B5,S%7D%0A17%20H%200%20%7B6,S%7D%0A18%20H%200%20%7B6,S%7D%0A19%20H%200%20%7B7,S%7D%0A20%20H%200%20%7B7,S%7D%0A21%20H%200%20%7B8,S%7D%0A22%20H%200%20%7B8,S%7D%0A23%20H%200%20%7B9,S%7D%0A24%20H%200%20%7B10,S%7D%0A
In fact, in RMG-Py, we are not even getting reproducible thermo. If you refresh the page above, sometimes we get cyclopentane and cyclopentene ring corrections, other times it will result in a Exo-tricyclo[5.2.1.0(2.6)]decene correction. This occurs by chance, resulting in completely different thermo.
