The preliminary accident report says the plane impacted in a horizontal attitude, at 180 degrees from the flight path to the runway. The Flight Data Recorders show advanced thrust (throttle) settings, the gear in the process of retracting, the stick shaker and stick pushers both operating, and the stall warning sounding.
My guess is that they were iced up and started to lose lift at around 5,000 feet until they reached stall conditions, at which point the stick shaker and stall warning operated, and the stick pusher automatically tried to pitch the nose down to lessen the angle of attack of the wings (excessive angle of attack, not inadequate airspeed alone, causes stalls). You can hear the stick pusher warning on the doomed Air Florida Flight 90 that crashed into the Potomac in D.C. in 1982. In the event the stick pusher inadvertently activates, pilots can either overcome it by pulling back on the yoke or disabling it.
This is pure speculation, but I should not be surprised that the instinctve reaction by the pilots to the stall warning and stick pusher activation so close to the ground was to go to maximum thrust and pull back on the yokes to override the stick pusher, ending up in an irrecoverable flat spin and riding the plane down into the house.
Angle of attack: what's that? Skip the next bit if you're not into technicalities of flight. Angle of attack (AOA) may be covered in flight training, but I imagine usually inadequately. I learned in the RAF, and it was drummed into student pilots what AOA is and why it is important to understand how it affects flight. I still have my classroom notebook.
AOA is the angle between the chord of the airfoil (the width, measured front to back) and the relative wind over the wing. Lift is derived from wing area, the airfoil (the shape of the wing, looked at from the end) and AOA. Increase the AOA and you increase lift (which is why pulling back on the stick, or yoke, usually makes the aircraft climb), so long as the airflow does not separate from the wing’s surface. If it does, due to excessive AOA, lift instead decreases, causing a stall.
My class notes have this, underlined twice: A wing can stall at any power setting, airspeed or altitude, banked or level, climbing or not, loaded or light. Angle of Attack is the deciding factor.
Anyone who has gone through primary flight training has almost certainly had his/her instructor saying - with greater or less volume depending on how close to the ground - "Watch your airspeed!" Very few, I imagine, will have been told to "watch your AOA," mainly because, again I speculate, civilian general aviation instructors have only a basic idea of AOA, and because in any case the airspeed indicator (ASI) would have been the only instrument on the panel to give you any indication as to the lift on the wings. The problem is that ASIs are notorious for information lag (the seconds it takes for a change in speed to show on the dial), only give one parameter affecting lift, and as airspeed decreases, usually when beginning to land, ASI error increases just when you need it most.
Anyone who has done glider flying will be familiar with the rudimentary turn and slip indicator on the trainer - a piece of string on a bit of rod sticking up in the nose. Keep the string streaming back directly at the windshield and you are properly coordinating the stick and rudder in a turn. It was the only "instrument" the Wright brothers had, and, although of course the term was not in use at the time, it also served as their AOA indicator. If it started to droop they knew a stall was coming. So, AOA indicators of one kind or another have been with us since the dawn of manned flight.
Most airliners have AOA indicators installed, but often only connected to the autopilot; not much help if you're landing a plane manually, as the pilots obviously were. In other words, unless they had an AOA indicator in their panel, they would have no idea if it was excessive AOA leading to a stall.
To demonstrate the importance of an AOA indicator, the US Navy and Marines have "flown AOA" in carrier landings since 1957, ignoring airspeed itself, and cut their accident rate by half the first year they adopted this procedure.
Again, this is just speculation. Perhaps the pilots had AOA indicators, fully understood their use, and knew exactly what was happening, but were handed too many factors in too short a time to be able to react and apply the necessary correction. Maybe the plane was so iced up that nothing they could have done would have saved them once they dropped below the commitment altitude.