holds_alternative<T>, get<T> overconstrained

[pdimov]

There's no good reason for these to not compile when T occurs more than once in the list of alternatives, as it's possible to make them work as expected. In the unlikely case the static_assert is desired, the user can easily write it himself.

[pdimov]

The counterargument here is that in cases like variant<time_t, ptrdiff_t>, sometimes these can end up being the same type after porting the program to a different environment, and get<time_t> will silently succeed when code has (tried to) put a ptrdiff_t into the variant. At present, get<time_t> would fail to compile after the port, which would catch the problem.