Anyone using a percentage past the first month isn't using logic.
We MIGHT see a max a 75M Difficulty increase every month (Obviously not the first couple months until BFL Bitfury and Avalon get their act together.)
That image shows the REAL increase for the month up to date 75%. What new hashpower leapt onto the scene to make that happen compared to the stuff we all know is about to come online in the next couple of months (and possibly some we know nothing about which are private projects)?
Granted this kind of increase isn't likely to be indefinite, but as KnC plan a next gen in march and prices drop and other will be planning new gen stuff too...it's not looking to me that it's going to slow in the next 6 months at least.
Look at the figures in that image for the projected difficulty in March and onwards. Massive.
Logic is exactly what is being used. Rather than wild optimism. Use the most reliable figures you know to estimate the future performance, the shelf life of the current batch of KnC miners isn't long at all.
What would be good to know is how accurate that calculator has been with something actually hashing, like a BFL Jally...given historical difficulty figures. I can't find anyone who's done that
I think that we're close to the point where the risk of buying a pre-order rig that may at best just break even, or an off the shelf rig that's already struggling vs just buying BTC and knowing you'll get at least what you paid for....that may result in less sales and a rise in coin prices to make mining worthwhile for those who can afford it and nudge the rigs struggling into a better return in the future. A rise in buying of BTC that drove the price to 130 dollars...that would help a lot of people to break even who might not currently?