Daniel, thank you for sharing all of this. I find it incredibly insightful....job well done!
One big question though....Why don't you build hard stops into the scripts? It seems that you've built this amazing machine to eliminate (or at least greatly reduce) human processing/emotion/error and allow the machine to calculate the risk down to .05 increments. Why not let it execute on that calculation also instead of risking human emotion/error and missing the manual stop?