Almost every crash is a first of it's kind. If you applyed your logic to these below and many others then these accidents wouldn't of happened the way they actually did. Anyone else care to add other flights?
US Air flight 1549: Dual engine failure by birds.
American flight 587: Rudder comes off.
United flight 232: Catastrophic hydraulic failures.
Alaska Air flight 261: Jackscrew in tail.
No, almost every crash is something that has already happened many times previously. The accidents that you cited are the exceptions. There is a proven scientific principle that the simplest explanation is most often the correct one. In this case, no one was reporting any unusual icing, the airplane departed controlled flight with pitch and roll excursions, and impacted the ground in such a way as to indicate a possible spin. People want to believe that this was something unusual because they find it difficult to believe that a 2 man crew could let the airplane stall. It has happened before. I am also not saying that icing had no part in this.
And with a tail stall you don't have the roll gyrations because the wing is still flying so you still have roll control. At least that is my understanding.