Why might it be beneficial to differentiate between causal-loop and sequential logic when setting project objectives?