make sure that the rationale behind #209 changes makes sense, otherwise work on a fix

[FerriolCalvet]

#209 has been merged after checking that the number of synonymous mutations estimated by the globalloc approach works approximately well, but revise if the current implementation is correct, and then see if any changes are required