Not sure what their motivation is, but if they want to keep data separate, they can load it into the same Application Group, with different Applications, and use the query restriction to keep different groups from seeing other branches documents. (ie, Branch ABC123 has a "where Branch like 'ABC123'" restriction in the group, and different groups can only see their own branches).
But yes, if they want the reports in different Application Groups, then they need separate spool files.
-JD.