Except of course for the fact that the idea that humans only use 10% of thier brain function is an idiotic misconception. And equally, postulating that olympic atheletes are at 30% of human physical potential has no basis in fact, evidence, or even simple logic.
In 1896, Spiridon Louis won the Olympic marathon with a time of 2:58:50. In 2008, Samuel Kamau Wanjiru won the Olympic marathon with a time of 2:06:32. So, in 112 years, with all the medical, fitness, clothing, hydration, training etc. advances, there has been a 52 minute improvement. And for the most part, this had been a very slowly creeping progression, with the improvements becoming lesser and lesser each year.
Now, if we are to assume that the Olympians who run the marathon are at just 30% of peak physical fitness, then you are in effect saying that you believe a perfect human could run a marathon in about 0:37:57. Or in other words, run for 42km non stop at a speed of almost 67kmph. Now, it doesn't take a great leap of faith to realise that this will never be naturally humanly possible. But to put it into perspective; record-setting sprinters Michael Johnson and Donovan Bailey hit top speeds of 43kmph, but only for seconds.
Below, too, I have posted a link to a mathematical model for the theoretical fastest humanly possible 100m sprint time. It is estimated to be about 9.2 seconds and is likely not to be achieved for hundereds of years. Usain Bolt already runs a sub 9.6 100m. If this was 30% of peak human capability, then Cap would be running the 100 meters in 2.9 seconds... or at 124kmph.
http://scienceblogs.com/startswithabang/2009/08/the_math_of_the_fastest_human.php