I ran some back of the napkin predictions for us. Figured the rest of us might be curious. If I overlooked anything or messed up in Excel let me know. I can also share this spreadsheet if you'd like.
I used bitcoinx to make these predictions. Also note, the actual return will probably be less given the pool/block finds.
I chose to use friedcat's 200TH prediction for the network rate by the end of 2013. Personally, I think this is exaggerated. To support that I believe it is in the best interest of miner's to use "worst case" predictions as a standard. It would be absurd to present an astronomical and incorrect number. It should, however, remain in the domain of possibility while at the same time remain large enough to dissuade the little guy. I also expect the hash rate to level out - or at least the market share. It's my prediction that eventually mining will become so competitive only the previously established mining co-ops will be able to regularly upgrade their equipment and stay in the game.
I would be interested in running this again with the Hash rate predictions from what BFL, Avalon, ASIC and KnC (possibly TerraHash?) are going to be bringing to the game in the next year. But for now I think we at least have an insight.