OK, one last post.
Although I've been siding with malmo, et al, this entire debate, it's not because I don't think that you could very well be right about the whole forefoot thing. There may be a 1-2% advantage, but to aver it as so, at this juncture is just hypothesizing. Go out and do tests, show it; that's what science is about. There are so many 1-2% effects in running that to look at this one very mild correlation that exists among the tip of the top of the current crop of elite runners that is also so heavily correlated with many other aspects of humanity on Earth (read East Africans) is just making a meritless observation.
I have not found any well-grounded physics justification to one of the common assertions that forefoot striking is better because it returns more energy for locomotion. For example, a guy I know at Wisconsin looked at a toy model to determine if there was an optimal amount of bounce in a running stride. Here's a link:
http://uw.physics.wisc.edu/~reardon/Central_Force_Motion.pdf
Lotta math to wade through, I know, but the upshot seems that optimum leg stiffness and "bobble" could not be determined from a simple physics point of view.
A more realistic attempt that I have seen to answer the question objectively relies on a more body-realistic computer simulation rather than physics/mathematical toy model. Their objective is to be able to answer biomechanical questions with simulated locomotion, thereby removing all those pesky time-variant variables that can plague real human testing. One thing they've tried to answer is the question of whether it makes sense to try to get hobby-joggers to forefoot strike. It's not published in a refereed journal, I know, but they concluded that heel-striking is actually more energy efficient for the pace (about 6:40/mile) at which they ran their simulated runner; not sure it's what they expected. Here's the link:
http://www.asbweb.org/conferences/2009/922.pdf
I don't claim to know how well this research is conducted, or where they plan to go with it. But it seems like a reasonable approach and I would be curious to know if others have knowledge of their work. It does suggest that the body can find many viable solutions to the running effort minimization problem, which probably also doesn't surprise me.
So this'll be my last comment on this thread (unless someone posts something academically interesting that's worthy of engagement):
Do what you think is right, believe what you want to believe, but just don't try to pass off speculation as fact.
Cheers