For any pilot who is landing an airplane, the goal is to have the wings producing lift right to the moment that the weight of the aircraft settles onto the landing gear. After that moment you want no more lift from the wings. That is why they design spoilers on the top of the wings. Immediately after landing you will notice the pilot activate the spoilers (flap-like panels that pop up to disrupt airflow over the top of the wings) thus keeping the wings from producing lift.
Airspeed management is especially critical during landing approaches. When transitioning from cruising speed to landing speed the configuration of the aircraft changes greatly as the airspeed bleeds off during the approach using things like flaps, landing gear, engine power, and adjustments in the pitch axis. Ideally you want the wheels to touchdown just above the the speed where the wings would stall (ie, stop producing lift). Then as the friction with the ground reduces the airspeed further (along with the spoilers used on complex aircraft) it becomes extremely unlikely that the wings could produce enough lift for the aircraft to become airborne once again.
The problem with landing in extreme gusty winds is how it affect the pilot's airspeed management. The amount of lift being produced changes many times, just seconds apart. Thus the pilot will increase the approach speed slightly to maintain a higher margin of controllability. But any extra airspeed that you are carrying as you bring the main landing gear to the surface must be reduced as quickly as possible because the wings are still producing enough lift for the aircraft to fly.
From what I gathered from the initial information, the RJ was landing into a right quartering headwind (not unusual by any means). But due to the gusty conditions the right wing would have been experiencing rapidly changing relative wind across the wing surfaces including the flaps. If you assume the pilot was approaching at a slightly higher (faster) landing speed (for the controllability factor) and couple that with the possibility that a wind gust across the right wing (the wing most affected by the crosswind), then it's very likely that a gust in the 40-50mph range could have produced enough lift as they touched down to roll the airplane severely to the left.
Naturally the first thing to impact the ground at that moment would have been the left wing. And with the roll being so violent, it would have sheered that wing near the fuselage inducing and even faster roll from the right wing which would have (likely) had its flaps fully deployed thus producing the greatest lift. Once it rolled past 180°, the right wing would have hit the ground sharply and created a similar shearing of that wing.
Of course this is just early guesswork based on some known landing procedures. But it is also possible that the airplane could have been approaching at a critically low airspeed and made it more susceptible to controllability issues when it encountered sudden strong gusty winds right near the surface of the runway. But either way it appears that the wings sheared due to a violent roll on or near the surface of the runway.
To me it seems inevitable that the gusty weather conditions will be a factor in the loss of control by the pilots. And if so, the issue of airspeed management will be thoroughly investigated. The 'black box' will give all the details of how that approach was managed in those weather conditions. It will be interesting to learn what the NTSB discovers from this investigation.