A tool comparison should separate time saved from errors not noticed. Faster only matters when review cost does not move downstream.