Previous commits went back and forth a bit on whether the various APIs
should use bytes or strings, but bytes appears to be a better answer,
because actual data in streams will always be 7-bit ASCII or raw
binary.  There's no reason to apply the performance penalty of
constantly converting between bytes and strings.

One drawback now is that lots of code now has to have "b" prefixes on
strings, especially in tests, which inflates this commit quite a bit.

This ensures that the same Python sub-instance handles the request,
even if it's coming in from two different virtual hosts.

Still needs work.

This is maybe 2.5-3 times faster than the list-based code, which
still isn't amazing, but is decent.