It seems like the natural next step would be "automatically deploy countermeasures while you still have [some threshold] remaining, then revert to manual below that," but I can also see how needing to train pilots in two or more different modes starts to pile up complexity costs and maybe leads to more mistakes compared to a simpler system.