If you were creating a new general purpose language language today would you target WASM? And what's the latest on garbage collection for WASM?
The big question is who will make the first WASM stdlib iface so backends can implement it. Otherwise, you might leverage LLVM right now because you have libc.