Hyper-DERP: C++/io_uring DERP relay - Same throughput as Tailscale's derper, half the cores

A developer built Hyper-DERP, a C++/io_uring DERP relay that achieves Tailscale's derper throughput while using half the CPU cores. Using kernel-level io_uring instead of Go's userspace processing, it eliminates scheduler overhead and context switches. Benchmarks show matching performance on 8 vCPUs vs Tailscale's 16 vCPUs, with significantly better packet loss and tail latency.