A better method of Data Replication for deletes

The current data replication process for deletes creates a sydatrep distribution record for every active location for every detail record when something like a price list for example is deleted.  It would be nice to see a method where perhaps just the header could replicate and that would trigger the POS to clean up the rest of the data on its own, or for deletes to be handled in such a way that they can respect the distribution filters (i.e. if you distribute out inprclst by locationid then only send the deletes to those locationids as well). I'm sure by using a delete staging table to then feed the data rep distribution something like that could be achieved.  Example the other day a couple of price lists were deleted, each had about 570 rows in the inprclst table for each pricelistid. This caused 570 X 2 rows to send to about 150 active DR locations which created over 17,000 rows in sydatrep in the middle of the day and backed up our replication of more important data (like dispatch vehicle assignments) by an hour. We also run into this in a lesser extent with dispatch order deletes when all the dsload records go out to all the distribution locations.