Originally Posted by
Gegenbeispiel
2. At the moment of blade departure, there will be a significant pressure drop (as before, when the blade was attached) from the back to the front of the blade. Thus, I'd expect the blade's departure path to somewhat behind tangential, opposite the direction of rotation and somewhat towards the nose. That's likely how the blade got around the cowl leading edge to smash the window.
There is no evidence that it was the blade that smashed the window - you are jumping to assumptions.