Both unitigger and BOG compute an a-stat. There is repeated code between both. This value is used by cgw but not by unitigger. BOG uses the arrival rate for unitig breaking.
Split the code into a common library that can be used by BOG and can run on its own. This way, adjusting genome size or a-stat options won't require re-running untigger. It will also improve consistency/readability between untigger and BOG and help the move to binary stores from message files.