This is something the compiler can't optimize very well. For one thing, the spec requires that they appear in the given order.

If this wasn't the case, a significant amount of code would break. I've seen a good bit of code where there's a struct A, and then struct B begins with the same fields as in struct A. This allows you to treat either of them as a struct A, a kind of polymorphism. The compiler doesn't know whether you're going to do that, so it can't arbitrarily rearrange the members of the structs.