About half of those are standard Rosetta scorefunction terms (van der waals attraction, dunbrack energy, vdw repulsion, lazaridis-karplus solvation, hacky electrostatics, hydrogen bonding terms split by donor/acceptor, then interchain_ is not part of the score function).
All are useful measures for final model selection. Their weights in the scorefunction (or just looking at the total weighted score) gives you an idea of what to do with each. For example, a model with mostly superb scores but a really bad fa_rep score means it’s got a bad clash somewhere and thus is probably wrong.
The 3.1 version of docking supports MPI. You need to make N directories (for N MPI processors) it to place results in before you start the run. I forget the expected directory names; I think the error message tells you what it wants but if you try it and can’t figure it out ask here again.
.
