Last time we saw two situations where you wouldn't bother trying to improve your forecast: When forecast accuracy is "good enough" and is not constraining organizational performance. When the costs and consequences of a less-than-perfect forecast are low. (Another situation was brought to my attention by Sean Schubert of