mdrun reports incorrect thread count use
mdrun has been reporting
Using 0 OpenMP thread per tMPI thread for PME
When it should in fact be reporting a >1 PP thread count. Some logic got obviously messed up.
Fixed parallel distribution and thread count reporting
The domain decomposition no longer prints the atom count for all ranks
(which gets far too long at high parallelization), but rather av,
stddev, min and max.
Corrected the OpenMP thread count print, which often printed 0 thread
for (non-existent) PME ranks.
Fixes #1681, #1685.