AssemblyScript / WASM findings

[donmccurdy]

Notes to self —

• Prefer StaticArray
• Use unchecked array read/write (180 MB/s → 680 MB/s)
• Any size optimization can have a performance cost, measure it
• Incremental GC has some overhead, measure it
• While the 'raw' bindings could be adapted for npm packaging more easily than writing custom bindings as I've done here, I don't yet see how to mutate the input+output arrays in place, without custom bindings.