We have spent many hours working on a better solution that can drastically improve the dupe finder, however certain issues came to light that would make it almost impossible to be a solid function. I won't get into it right now, but we are still thinking of other, better ways to tackle this issue.