so what is the break even point on $/kwh??
Even if hardware costs kept constant (which they don't--the cost of hashing power is continually changing and hardware also keeps depreciating, so factor that into your costs), that number is constantly changing because bitcoin mining is self-correcting: only 50 coins (soon to be 25 coins, starting in December 2012) are produced each 10 minutes, no matter how much hashing power is in the system. Imagine a magic goose that lays 50 golden eggs per day. No matter how many people are lining up to get eggs, there will never be more than 50 eggs per day. The more people crowding the goose, the less each person gets. Likewise, it doesn't matter if there are only 10 miners, each with a HD5770, or 1000000000000 miners, each with a pair of 5970s; the rate of generating bitcoins is the same: 50 coins per 10 minutes, meaning 7200 coins per day.
The earlier you got in, the less competition there was, and the easier it was to mine coins. (The first guy to mine coins had a monopoly and got 50 coins every 10 minutes!) But now, with people encouraging ever-increasing numbers of people to buy ever-increasing powerful GPUs and specially-built FPGA/ASIC hardware, the difficulty is sky-high and going higher.
People encouraging others to mine = higher difficulty. People buying and using FPGAs and special-purpose devices built specifically for bitcoin mining (ASICs) = higher difficulty. A single ASIC mini-rig can do the work of hundreds or thousands of GPUs. I don't think many people realize just how dramatic difficulty will change with the addition of just a few ASIC mini rigs. Imagine difficulty quadrupling overnight. That would be the effect of the first batch of ASICs being shipped out and turned on. The second, third, and subsequent batches will drive up difficulty even more (I've seen multiple reasonable estimates of ~20x difficulty, yes TWENTY times higher difficulty after the first month of ASIC shipments, and even higher after the second month of ASIC shipments) and eventually make it impossible for GPUs to compete, just like how GPU mining has obsoleted CPU mining because GPU mining is so much more efficient.
http://bitcoin.sipa.be/speed-lin-ever.png
Some fools think that mining difficulty drives price, thinking that if difficulty rises, so will the dollar-to-bitcoin exchange rate. That isn't true. Bitcoin has been around for years, and the total number of coins already mined is over 8 million. Each day's addition of 7200 coins (soon to be 3600 coins) is negligible. What really drives price at this point is demand.
Demand is volatile. If Silk Road (an online gray/black market using bitcoins as currency) goes down, for instance, that will have waaaay more impact on prices than some latecomer miners' behaviors. If the FBI decides to attack the choke points of bitcoin, that may choke demand even more, because it'd be harder to convert dollars into coins. (Bitcoin is NOT as impervious to attack as some fanatical true believers think it is; you can still attack exchange points, like banks leaning on Paxum. Imagine a situation where Dwolla refuses to take part in any dollar-BTC exchanges anymore, and nobody else steps up either because the FBI and DOJ threaten them with shutdown. MtGox is beyond the reach of the US Govt, but Dwolla and all Dwolla-like entities are vulnerable to government interference.)
Even if we assume that bitcoin actually survives for another year, what happens when more people realize that bitcoin is not good for commerce due to the lengthy verification time? Unless there is a dominant, secure, trustworthy instant-credit-debit bitcoin company out there, 10 minutes or more is a long time. A bitcoin-like system that acts, say, 100 times more quickly, would be more useful for actual commerce.
Supply is also volatile but less so due to the fact that over 8 million coins have already been mined. And if supply on the market does change, it's due to old timer miners, not newcomer miners. There are a lot of miners and bitcoin enthusiasts out there who mined for a lot longer, a lot earlier than you did, or who bought bitcoins early on for pennies per coin. When they decide to cash out thousands of coins at a time, that moves the market way more than any of the latecomer miners.
Btw, if you are so sure that bitcoin prices will go up, then it's more profitable to invest directly into bitcoins themselves and not buy mining hardware. In real life, who makes more profit per dollar invested: timber/mining/hardware companies, or investment banks/private equity funds/hedge funds/etc.? Same thing with bitcoins: smaller, less risky profits to miners (who can sell their equipment if bitcoin crashes, except that FPGAs/ASICs may have little value), and larger, riskier profits to those who purchase bitcoins directly.
TL;DR version:
Don't think that difficulty will remain constant when doing your profitability forecasts, especially since mining difficulty has increased for the last several months and will likely keep increasing as more and more people mine with faster and faster hardware. Once purpose-built ASIC miners come out (as early as October), that will increase difficulty by a factor of 10-20 in the first few weeks alone, and even more after that. Plus, bitcoin generation halves in December 2012. And for every person you encourage to mine, you decrease your own mining profits, especially if that person then tells 10 other people, who then tell 10 other people.... (Despite this, you guys keep bumping this thread up anyway, lol.) And don't expect price to keep going up; price may drive mining investment to some degree due to the lunatics out there willing to gamble big money on something as risky as bitcoin mining, but mining difficulty does not drive price.