The code is already multithreaded. The main thread issue is not going to go away with DX12, since this is the thread that syncs up the different workloads offloaded to other cores/threads.
I’m hoping they’ll be able to make the WASM pipeline multithreaded though, since that is currently a very limiting factor.