Yes, the sunsets changed, they are more yellow with WB, but I'm sure that this part is not entirely right. During the sunsets the contribution of sun to the computation of WB is minimal, because the angle of sun rays is almost parallel with respect to the measuring surface. But as I think about it more, our brains aren't constrained to perceiving white on horizontal surfaces, and something has to be done about it.
It seems to be right up until the sun approaches the horizon, while the sky is still lighted well, but then it needs to be adjusted, maybe slowly suppressing the WB effect.
Also, the sun disc isn't right size, should get bigger and possibly deformed near the horizon as the optical depth rises.
