The profit function keeps increasing after the last price point. How should I estimate the optimal price?

Unfortunately, you cannot.

The Gabor-Granger method, and the specific response function implemented in Enginius, is pretty good at interpolating between price points, but cannot be used reliably to extrapolate outside the tested price range.

If the optimal price solution seems to be above the highest price point tested in the survey, the best strategy is to collect another data set with a wider price range.