When materially new evidence arrives after something has already been assessed, the earlier conclusion has to be re-qualified before it guides the next action. Two things are read separately: how the evidence relates to the earlier basis, and how well established the evidence itself is. Then each action is read against both.
A change in evidence does not settle a new conclusion by itself. It re-opens the earlier one — and where things land depends on what the evidence touches, how well established it is, and what you're about to do.
The new evidence has two independent properties: what it relates to, and how well established it is. A direct-but-unconfirmed report and a direct-and-confirmed finding are not the same input. Set both, then read the action.
Why the two properties are read separately. "Directly affects the basis" is about what the evidence is about; "independently confirmed" is about whether the evidence can be relied on. A direct report that is still unconfirmed is enough to stop the earlier assessment from automatically governing — and enough to justify reopening and preserving — but not enough, on its own, to justify an irreversible external action. Only when direct evidence is independently confirmed does the earlier assessment become no longer eligible to govern at all.