I think we can all agree that hash rate will eventually level off at a certain point, based on the most efficient mining chip available at the time and what is considered to be an adequate return. Based on the charts in the links below, I think it would be foolish to think it will level off within the next 3 months as many new chips are being delivered to many customers.
http://mining.thegenesisblock.com/https://blockchain.info/charts/hash-rateIt makes sense that eventually the increase has to return to a linear function as only so many mining chips can be produced in a week. I think we will hit 100,000 TH/sec sometime in May before the leveling off will occur.
We do have actual data that we can now use to predict the amount we should earn. Consider the following stats taken from the PETA page on Eligius:
Estimated Earnings:
Your approximate maximum potential earnings at the current network difficulty of 3,129,573,174.52 and maintaining your 3-hour average hash rate of 52.14 Th/s is 8.37829676 BTC per day.To make things simple, currently the pool tells us PETA will mine 8 BTC per day at 50 TH/sec. This means at todays difficulty a 700 TH/sec mine would churn out 112 BTC a day or 1344 BTC every 12 days for that difficulty period. Based on the B.MINE shares, difficulty has been increasing 20% every difficulty period. This means that mining payout would be 0.83333 (reciprocal of 1.2) of what it was the prior 12 day period. This Friday's amount will only be around 64 BTC. This Friday is coincidentally when the next difficulty increase will occur.
Feb 28 64 BTC (1344 if we had been at full power (700 TH/sec) for the whole 12 day difficulty period)
Mar 12 1120 BTC
Mar 24 933 BTC
Apr 05 778 BTC
Apr 17 648 BTC
Apr 29 540 BTC
May 11 450 BTC
May 23 375 BTC
Jun 04 312 BTC
All the above numbers assume a 20% difficulty increase each period (12 days), and also each difficulty change will occur 12 days apart. Keep in mind that we will not be at full power till the end of May. The above numbers from Feb 28 till Jun 04 sum to 5220 BTC. Since we are not at full power till the end of May, I would say we will mine half of that which is 2610 BTC. Also accounting for reinvestment we would end up with 2610 * 0.65 = 1696.5 BTC for dividends. 1696.5 / 100,000 shares = 0.016965 BTC per share. If you bought a share at 0.05 you will have paid for roughly 1/3 of it by the beginning of June.
Since mine is fully deployed at the beginning of June, let's assume that the 35% reinvestment completely offsets any further increases in hash power (difficulty). This means PETA should be able to keep mining 312 BTC every 12 days (182 BTC a week) till the next block reward halving. This means that 118.3 BTC (182 * 0.65) should be available for dividends each week from June 4 afterwards; this is 0.001183 BTC per share. Based on an initial value of 0.05 this amounts to a dividend of 100% per year. I know CryptX also has hosting fees which I have not accounted for, so they will affect the above numbers.
Incidentally in Alberta, Canada I pay $0.08 per Kilowatt hour, so maybe consider moving the mine here. The whole province sits on top of coal deposits, so we can keep our coal power plants running cheaply for centuries to come.