As Hull said, "if rho=1, and sigmaS=sigmaF, than h*=1", which is just the solution to the example you pointed out. Remember, hedge ratio h* is not contract ratio. And pay attention to the concept of rolling the hedge.
Why "理论假设虽然设定其均值是0"? Sigma is sort of standard deviation, it must be far from 0, even using simple estimation. In practice, people use models like GARCH to estimate sigmas.
In Hull's book, the objective of hedging is to minimize the risk the the hedging portfolio, so the expectation of price change has nothing to do with hedge ratio. You may double check the derivation.