Missing evidence should remain visible. An unknown result identifies a review task; it must not be presented as a passed control.