No one ever said it would. Remember I used only 7% per week in my simulations. Even in the  very unlikely event the next difficulty would drop a little bit (Im willing to bet it wont), we are still way above the assumptions I used in my projections because of the previous 25% increase. Also keep in mind that an early big rise followed by a slower rise (or even a small drop) is far worse than a more constant rise.