Right, and I can construct theoretical models where this works well!
But they're all ripe for abuse, and that makes them politically untenable.
As a layperson in such social matters, I'd guess the real-world abuse rate would be something like 10%. This is mathematically tolerable, and socially beneficial.
But even at 0% real-world abuse, there would be righteous opposition. And at 0% + 1 predictable example, it'd be politically toxic in most of the country.
