If the difficulty goes up too quickly, then you reinvest and reinvest, but the amount of BTC you earn doesn't go up very much at all, you just use more and more electricity.
So really there is no way to know.
Yeah I see what you mean. Ive done a bit of math and it seems to work out up to 100m difficulty. Probably a lot more but only did 100m thus far.
Im pretty dumb but it seems to me if I buy with USD 20 usb asics to start with that alone will recoup costs within 300ish days at current difficulty. Of course by then it would be so high they would not end up paying for themselves. HOWEVER if after say 2 weeks I have enough to buy another usb miner, this time with just bitcoin. Now if I compared/counted the exchange rate for USD to bitcoin it will still not be profitable. However consider I am in essence creating money out of thin air. So why even bother compare it at this point.
So the first two weeks after getting the 20 miners I can add one more. Then even less time after that I can add another, and so forth and so on. Basically doing all this at only the cost of electricity and USB hubs which in my area is dirt cheap. 0.09 kw/h for electric and $25 for a hub. At the point at which I double to 40 miners, thus reducing the orginal cost of miners break even point in half, or maybe a quarter at worst assuming difficulty jumps super sky high within 1-2 months.
Bascially I dont see how this isnt profitable.
At the very least I will end up mining enough BTC to buy a better/more efficient miner and replacing the costs of the original 20 miners.