How forecasting progress, using burn-downs, burn-ups or cumulative flows, might be useful for the team and the organisation? Does it supplement or conflict with empiricism?