Yes, though the loop unrolling is news to me. I'll have to keep
that in mind next time I'm trying to squeeze some extra
performance out of a loop.

btw, found a static switch enhancement request here:
https://issues.dlang.org/show_bug.cgi?id=6921