I’m disposed to liken it to an anytime algorithm (https://en.m.wikipedia.org/wiki/Anytime_algorithm). You want your projects to be as ‘anytime’ as possible, IMHOBOLHE. I think that’s what your comment is driving at: that there’s a significant benefit to a plan that delivers half the benefits if it only runs for half the (eventual, real; not planned) time, vs a plan that requires a long investment of time but only - hopefully - pays off in full right at the end.