Test Outcome

• An empty update loop made with a coroutine is roughly 3x slower than an empty Monobehaviour update loop
• The amount of garbage generated by starting a coroutine seems to be independant from the size of its class or the amount of things referenced inside the coroutine.