I applaud the stricter rules in general, as Mark noted nicely, the issue is that `__index__` is maybe a strange way to achieve that for bools (it is not like `123` is a clean bool)? `__nonzero__` coerces to bools, there is no `__bool__` to convert to bool safely.

Basically: this seems to force numpy to back down from saying that `list[np.True_]` will be invalid in the future. (And we cannot just get rid of our bools unfortunately).