Instead of maximizing “benefit - cost” I would recommend maximizing “cost/benefit”. The former will bias towards big tasks with big payoffs, but often there ar

Instead of maximizing “benefit - cost” I would recommend maximizing “cost/benefit”. The former will bias towards big tasks with big payoffs, but often there are small tasks with extremely good ROI, which will get lost if you just look at the magnitude of the net benefit. But, +1 to the general approach of trying to put a number on it. Remember though, your numbers are uncertain, beware the McNamara fallacy.