One thing you guys haven't considered is that you are chasing your tail when you put a real pole in front of something moving. Here's the problem:
1. When you hit the pole, it deflects, energy is absorbed and a load is applied.
2. The larger the pole, the higher the load. The energy stays the same, however.
3. The larger the pole, the stronger the pole.
4. The problem is that if the first iteration doesn't work, if you use the same material and configuration (cantilever pole), no other iteration will work if you don't plastically deform either the pole or the object that hit it.
Try a few iterations - you'll see. What it boils down to is that deflection is inversely proportional to moment of inertia - and SO IS STRESS! What that means is, changing the moment of inertia (by making the pole bigger) decreases the deflection, thereby increasing the load just enough to stress the pole to the exact same value. Odd - but that's how it works.
The point is, for a given material and impact energy, you may have to go to another energy absorbing setup. Unless you consider plasticity, which is a big bag of worms...