But that number is not relevant. What is relevant is the ratio of humans avoiding these accidents to humans not avoiding these accidents.
Like I said, having to avoid the car in front which braked hard is routine. It happens all the time. It's why the blame and culpability for a rear-end accident is always on the rear driver by default - safe driving distance and all that.
Which is why I phrased my question to the GP the way I did - if he managed to go 300000km without rear-ending someone even though unpredictable braking happens so routinely ...
> What % reduction in these would make it make sense to "require" this?
I doubt there'll be a decrease, if any. A better system that beeps if a safe minimum following distance is not adhered to will almost certainly prevent more accidents than auto-braking, because the auto-braking system will sometimes get it wrong and cause the jackass tailing you to rear-end you.
An alert won't cause any accidents.
I don't have an auto-brake feature, only ridden with people who do. I believe that most of them do have some sort of warn prior to brake - but to your point, there are so many situations where this could falsely trigger that it would likely end up being tuned out, at least as I imagine it. Actually, my truck has one of the "warning" things, and I ultimately turned it off for this reason. The number of "immediate danger" warnings that would flash when I was driving, e.g. alone on a country road at night, made it not worthwhile to me. I can imagine an auto-brake would be even worse that way though, lol.
As an aside, trying to quickly search for these sorts of statistics leads to an awful lot of law firm websites.