It is a formal logical fallacy. But if we step out of the ivory tower and into the real world, the elites MUST be doing a good portion of their training correctly. Otherwise, they wouldn't be elite. There is a finite number of things that go into becoming the best athlete you can be--training, diet, sleep, altitude, etc--and training footwear, or lack thereof, is among these. Perhaps elites have 95% of their training "right," and are wrong on the shoes issue. But if that were the case, wouldn't we see more elite runners running barefoot? After all, it's been a long time since Abebe Bikila and Zola Pieterse (who both later wore shoes, mind you) were winning major competitions barefoot. And Christopher Koskei DIDN'T win in 1995 (WC steeplechase, for the uneducated)! As Renato Canova said, "It is better to be a champion than a personality." And right now, we are getting 17:30 5k runners, who are not champions, parading around like personalities because they train barefoot on pavement.
Let me reiterate: THERE IS NO SUCH THING AS A BAREFOOT SHOE. You can call your thin rubber skivvies "barefoot shoes," but they are fundamentally and kinematically different because they interrupt the interface between the ground and the plantar surface of the foot. Even 4mm of rubber completely changes the sensory feedback from the plantar mechanoreceptors. Not that I'd expect you to know what those are, much less how they relate to "feeling the ground," as you say. In fact, running in a smooth-bottomed rubber slipper on pavement is about the WORST thing that you could do. Is it any surprise that podiatrists are reporting a spike in metatarsal stress fractures from vibram fivefinger wearers? (http://www.podiatry-arena.com/podiatry-forum/showthread.php?t=48566)
Secondly, if you don't think Galen Rupp or Alan Webb would immediately ditch shoes if they had good evidence that training barefoot full-time would make them faster, you are a fool. Shoe contract or no shoe contract, an Olympic caliber runner will do anything in his power to get faster.
Thirdly, there is no evidence that running barefoot strengthens your feet. It is not an absurd proposition, and I agree that it probably does strengthen your feet, but this has not yet been borne out in scientific research. But furthermore, there is no evidence that stronger feet prevent injury. Again, I agree with you that stronger feet probably DO prevent injury--but it has not been shown yet. So please refrain from making such statements until you have a citation to prove it. And please don't cite Robbins 1987, because it is completely fraudulent and dishonest.
As I said before, running barefoot on grass is fine. I support it. I encourage it. I do it myself! However, running barefoot, or in thin rubber shoes without any cushioning, is irresponsible, as you are exposing your lower leg to a surface/footwear combination that is outside its zone of optimal stiffness. The body works to mediate this somewhat by increasing your stride frequency and increasing the load on the forefoot (to reduce peak pressure on the heel), but both of these can be accomplished while wearing shoes. Running barefoot on grass is fine because the grass has enough "give" to return energy and reduce the shock absorption needed from your legs.
Now, if you'd like to have a discussion about different parameters of shoe design, I'd love to have that talk. There are a LOT of things in modern running shoes that do not have much (or any) research supporting their incorporation into training shoes. Pronation control, big heel-to-toe drops, stiff TPU shanks, and complicated, multi-layer uppers do not have much literature support...and don't really even accomplish their purpose. The recent Nike study (http://bjsm.bmj.com/content/early/2010/06/26/bjsm.2009.069849.abstract) solidified the case against the "pronation control" paradigm. However, if you run in medially posted shoes with TPU bridging and a stiff heel counter and you don't get injured, more power to you!
I don't like the ASICS hyperspeed at all--the lasts ASICS use don't fit my foot right, and the upper is sloppy, too loose, and has too many overlays--but the midsole and outsole of the shoe are pretty close to the "ideal" running shoe. A young kid who is new to the sport and doesn't have any underlying foot/leg problems (equinus deformity, previous shin/ankle/foot injury, corrected clubfoot, etc) should probably get something like the hyperspeed--low heel-to-toe drop, a reasonable amount (not too much, not too little) of cushioning, no TPU plastic bridging, no dual-density EVA. No sense getting used to big bulky trainers if you don't have to. The pendulum of the industry seems to be swinging towards minimal shoes now--Saucony is dropping most of their performance shoes (A4, Fastwitch) to 4mm, and I hear they have a zero drop shoe coming out next year.
I'm pretty excited about the New Balance Minimus--it has all these characteristics, plus no heel counter, a wide forefoot, and very little rubber on the outsole. Nike has proven with the Free line that the new EVA foams are sufficiently durable to function as a midsole AND an outsole, so additional rubber is not really necessary on roads. Rocky, rugged trails are another issue, but that's not really my specialty. Of course, I haven't actually worn the Minimus yet, so maybe they botched the upper, made the cushioning too firm or too soft, or something else, and it'll be a piece of crap. So who knows.