Reverse generics and opaque result types

[orobio (Manolo van Ee)]

@CTMacUser

while the other two can't really have an all-the-constraints-are-in-the- where -clause equivalent without looking noticeably different from the versions given here with the constraint inline.

I agree that using a modifier for the typename itself results in something that is most consistent with existing syntax. However, I think it's best to discuss these alternatives in the existing proposal review thread: SE-0244: Opaque Result Types