Post-divides event / POI merge and hyREfactor

The biggest thing of note with this MR is that I short-circuited the "map_outlet_ids" function to include split events into the gf_outlets dataframe used by "create_run_subsets". I found sometimes that an event put into "map_outlet_ids" with the reconciled layer would yield the incorrect reconciled ID associated with the split event. The work-around seems to yield correct results, but have just tried it on HI (with AK next).