A 50/50 matter/anti-matter system could still house stable local pockets of mostly matter or anti-matter. The problem is, from what I understand, that the universe seems to have sprung into existence with way more matter than anti-matter, and we don't know why.
If you flip 2n fair coins, you expect n+δ heads and n-δ tails, where δ is (IIRC) sqrt(n/2). Going much away from that becomes infintessimally unlikely.
Probability is a subject famously easy to get wrong, so be careful with what I'm about to suggest: I *think* you could argue that in the moment prior to the inflation epoch spreading everything out just enough that pair production stops*, any given particle in our horizon is a coin toss of matter or antimatter.
Number of observed atoms in the universe is about 6e79 (http://www.wolframalpha.com/input/?i=how%20many%20atoms%20in...), so 6e79 = sqrt(n/2) -> n = 7.2e159 due to protons, and the same again for electrons; as we don't see significant signs of antimatter, any around must have annihilated a long time ago, so in this scenario we should expect to see ~7e159 (red-shifted) photons from the supermajority of particles which have annihilated.
It's outside my field to know how that compares to cosmologist's observations.
* won't that be at different times for protons/neutrons and electrons?
I can't get good answers on the expectations for either "why are protons and electrons counts the same" or "what is the observable consequence if they're not?"
