I have run some symmetric_docking jobs with cyclical symmetry in which the "ten lowest-energy structures" all look like either clockwise (CW) or counterclockwise (CCW) pinwheels (see jobs 17648-17650). In the Score/RMSD plot for each of these runs, there is one funnel with RMSD near 0 Angstroms and another funnel with RMSD near 26 Angstroms. I am thinking that all of these lowest-energy structures should really have RMSD values near 0 Angstroms. My guess is that for the 5-subunit jobs (17649-17650), you could align the CW and CCW pinwheels to all give RMSD near 0 by aligning the CW subunits A-E with the CCW subunits E-A, respectively. Similarly, to get RMSD near 0 for the 6-subunit job (17648), you could align the CW subunits A-F with the CCW subunits F-A, respectively. Does this make sense? If it works, perhaps future versions of symmetric_docking will automatically do this so the funnels with RMSD near 26 Angstroms would become funnels with RMSD near 0 Angstroms.
The rmsd is reported as a symmetric rmsd for up to (I believe) 6 subunits. This rmsd is calculated by creating all possible chain orders, evaluating the rmsd for each and taking the minimum. The number of chain orders increases as a factorial with the number of subunits, which gets very slow very fast. So for large number of subunits (above 6), the symmetric rmsd is not reported.