We would like to have control over the match settings for "similar" addresses in constituent updates.  We wish to retain an address as new if it has a lower similar score, but update or ignore the address when it has a high or exact similar score.  Historical addresses are important to retain for prospect research and other people finding processes.  

In order to retain the historical address and add the new address, the current similar score setting means that we are also retaining exact matches which muddies our data and creates unnecessary batch exceptions.