I can't help but
chime in here. What seems to be missing from the original w/u is the notion of
sample size. We've all heard the phrase '
large N.' If
these unlikely conditions only have to occur
once, what if the
soup is the size of a
planet? Take
Earth's oceans. If you divide up their
volume into spaces the size of these potential
proteins, then what happens?
There's also the notion of time. If we assume that a 'snapshot' of the soup state (God I miss my Newton) lasts for as long as it takes these potential protein spaces to completely slide past each other, then you have even more chances. If it takes a minute, even, for the soup to mix, how many minutes has the Earth been around?
The problem with these arguments, as The Alchemist has properly noted, is that they can be reduced or extended ad absurdum. It's all in the assumptions, baby.