I think reducing keystrokes is a proxy metric for reducing the time and cognitive load of accomplishing an editing task. It's easy to measure, and probably good

I think reducing keystrokes is a proxy metric for reducing the time and cognitive load of accomplishing an editing task. It's easy to measure, and probably good enough to tell if you're getting anywhere, qualitatively. I don't think it's good enough to quantify improvement in a meaningful manner, though.