Let me preface this by saying that I have no intention of changing any of my equipment at this time. This question is being asked more out of curiosity since random optimization problems are always on my mind. And perhaps it will help guide decisions on equipment at some point in the future.
Evidently, a shorter crank makes it easier to spin at high cadences. So does that mean a longer crank makes it easier to climb really steep stuff when you're running out of gears? Let's say you're on a steep climb, already in your lowest gear, and watching your cadence bog down. Presumably, since you get more leverage with a longer crank, you can generate more torque, and if you're out of gears and at a low cadence, more torque should equal more power. Argued from a slightly different angle, does a longer crank optimize for a slightly lower cadence, thus allowing you to take on a slightly steeper grade before your cadence starts bogging down?
So what do you think? For what it's worth, Sheldon Brown advocates calculating a
"gain ratio" that takes into account crank length, and based on the numbers, he seems to advocate the belief that a longer crank equals slightly lower gearing. And I know Lance Armstrong went to a shorter crank to improve spin, but unlike him, us mere mortals do run out of gears. Since he's not running out of gears, I don't think the analysis I suggest above really applies to him or any other pro.
So what about us mortals? If you have a bike dedicated to very hilly rides (or if you always ride hilly stuff), would a slightly longer crank make sense? Of course, it's no substitute for proper gearing, but would it help you get past the occasional 15-20% pitch (or whatever "borderline too steep" means to you) in the road?