What caught my eye was the oscillation in the abundancy graph, there is a tendency for even atomic number elements to be more abundant than odd atomic number elements. Looking at the original Wikipedia article leads to the Oddo-Harkins rule http://en.wikipedia.org/wiki/Oddo%E2%80%93Harkins_rule . There is more discussion on the stability of even vs odd atomic number elements here: http://en.wikipedia.org/wiki/Even_an…
when you roll 2 dice, you get a lot more evens than odds. i imagine the same principle holds. if an odd (Hydrogen) forms with another odd, you get even. Hydrogen+helium=odd, but helium + helium = even. as the evens outnumber the odds, even more evens are forming with evens.
odd + odd = even (odd reduced by 2, evens increased by 1)
odd + even = odd (evens reduced by 1)
even + odd = odd (evens reduced by 1)
even + even = even (evens reduced by 1!)
When the evens outnumber the odds, odd+odd will be rare, so the percentage of evens will tend to decrease until it reaches 50%.
No matter many of each type you start with, the equilibrium position is 50-50% odd and even.
However: there may be an exception to this. If all the odds are in one place, then odd-odd may be more likely than random. If you have all evens in one place, you can't get back to having odds again - you can't (ever) increase the number of odds that already exist to balance things out, you can only decrease the evens. 0-100% odd and even could also be an equilibrium position.