The hotplug branch in manuelnaranjo's repo has very broken support for
composite devices, blah. So let's switch back to libusbx master and deal
with that.

This is pointed to a bleeding-edge branch where Windows hotplug support is being developed

This is more secure by default, and it allows fcserver to work correctly on Mac OS when no networking is available. (With wifi disabled, a server bound to 'null' won't be listening on localhost. Blah.)

When staring directly at the LEDs, the dithering flicker is visible- but in an enclosure, the flicker is totally nonexistent and the linear section causes a noticeable discontinuity.

A pure nonlinear-only approach seems like the best default. I'll document cases when this linear section might help, and it will be up to individual clients (or fcserver config files) to enable it.

This is really dumb, we weren't processing backlogged packets from the opcsink buffer. This is the cause of the ugly delays I'd been seeing in the 'waves' demo.

This lets us disable the Nagle algorithm, and have finer control over error handling

This is a big improvement in dithering quality: Near zero, where we have the most problem with visible flicker, this uses a linear curve instead of the usual nonlinear curve.

Now the firmware config flags actually select between several different inner-loops. This will be useful in the future if we want to support different configurations with runtime config options, and right now it's useful for side-by-side comparisons with the various config options.