Because at scale, that doesn't work, I think, even money no object, because it's a task that requires a lot of domain knowledge, but also is monotonous and offers little in the way of satisfaction or looking good on your resume, so the candidates you're trying to hire aren't very junior, and also are not going to like the work.
Why not? "At scale" is not an argument on its own. This is something where scale helps you because the number of different hardware you need to support is independent from the size of your customer base.
And if you really can't act responsibly "at scale" then you shouldn't be allowed that scale.
IIRC, at least 15 years ago, that's exactly what they did. They published a tool suite that your driver should pass before submission because otherwise it would fail when they got to it. I think the submissions were fully automated by that time.