optimize is only knwon from common_mRM_mHM and therefore cannot be used
for the distribution of domains to the MPI processes already.
In case of sequential run, the master process nbow has nothing to do.