improve duplicate detection algorithm
When importing constituents even the slightest deviation in spelling, address format, using a nickname, etc, doesn't register as a duplicate when importing, only exact matches seem to filter. This means we are having to manually check for duplicat...