Do you think it would be practical to rewrite even the hard audio DSP stuff, like echo cancellation, in Rust? Or should webrtc-rs use bits and pieces of Google'

Do you think it would be practical to rewrite even the hard audio DSP stuff, like echo cancellation, in Rust? Or should webrtc-rs use bits and pieces of Google's WebRTC C++ library for that? I certainly see the benefits of not using the whole C++ library, with its threading model and all.