No breaking waves. Ever (unless researchers or I come up with a good idea): This method is based on heighfields so that means there is only one height per cell of surface. A breaking wave would have more than one height per cell.
Small waves generated when rotating a surface and an interactor together: Relative movement between a surface and an interactor is always taken into account. Translating both at the same time will not generate waves. However, if both are rotated together under the same parent a little wave generation cannot be avoided. Sadly, this is due to the lack of precission of the calculation of the angular velocity for both objects. A solution that would make this system less flexible would be to associate each interactor with each surface, and calculate the velocity of the interactor with respect to the surface, but this will not be done unless you have a good reason to share with me.
Open borders are still not perfect. Even when you don't fix the borders of your surface, a percentage of the wave will be returned. I'm still working on it.
An interactor with a kinematic rigidbody should have a trigger collider. But this is due to the native physics collision system. It the collider is not a trigger, it will not generate contacts, so there will be no waves.