That answer fails at larger alpha. For example, at alpha > 0.6pi, x>1.
Re: 2) It looks hard to calculate it exactly, the integral is very ugly. It is much easier if you're doing this on a periodically continued square, because then there are only two variables effectively and the answer is 0.38. This is a strict lower bound for your question (although obviously not a very good one).
The question also looks much more tractable if you use another norm, for example the max-norm or thetaxi-cab metric l_1 metric. Maybe that way it is possible to derive some other bounds on the answer.
Re: 2) It looks hard to calculate it exactly, the integral is very ugly. It is much easier if you're doing this on a periodically continued square, because then there are only two variables effectively and the answer is 0.38. This is a strict lower bound for your question (although obviously not a very good one).
The question also looks much more tractable if you use another norm, for example the max-norm or the