The find and replace tool currently does not run row by row, and finds anything in the find column, and replaces it with anything in the replace column. I was under the impression and designed my workflow to use this as a row by row find and replace, not entire columns.
A simple fix would be to allow users to group by RecordID, which should also speed up the find / replace tool for larger data sets I would imagine.
What I am going to do in the meantime is use Regex to replace the word out.
Thanks!