Triple is nothing. At the current price/GHs you would do a HUGE profit with a Jupiter if the difficulty only went x3 by the moment you get it.
As it's scheduled for September/October, I would consider a difficulty of at least 70 million, and then an increase of at least x100 in the following 12 months (and I wouldn't be surprised by x300...)
If then diff. is any lower than that, better for everybody, but running simulations with anything lower 70m in October and x100 increase in the following 12 months is just absurd wishful thinking IMO.