Test statistical correctness for no-between group contacts

This PR is to request the addition of a test for statistical correctness that addresses a special case in which the population has multiple demographic groups that have uniform within-group contacts, and no between-group contacts. Under these conditions the proportions of the final sizes overall should correspond to the proportion of each demographic group (with susceptibiity groups as modifiers). This is essentially a model of two separate populations with homogeneous contacts within them.