Severe icing is far from being proved even though some icing is suspected. A little bit of almost everything is suspected.
In a typical investigation the NTSB will look at the history of the aircraft itself and the history of its type or model.
In that regard there are some reports over the past few years that Q400's were having mechanical problems:
Scandinavian Airlines announced Sunday that it would abandon a fleet of 27 planes made by Bombardier of Canada that have been involved in crash landings.(Airline Stops Flying Q400 Turboprops). The Q400 line is evidently a newer type of aircraft:
The unusual step by SAS came after one of its Dash 8 Q400 commuter planes crash-landed on Saturday because of landing-gear failure, the third such incident involving the airline in the last two months.
The Q400 entered service in 2000 and can seat 68 to 78 passengers. It is the latest in a series of turboprops designed mainly for regional aircraft use that were first developed by de Havilland Canada, owned by Boeing before being acquired by Bombardier.(About 60 Bombardier Planes Grounded After Crash Landings). The main problem had to do with the landing gear, but there was talk of broader problems:
Both Transport Canada and Bombardier rejected suggestions that the series of crash landings could be a sign of broader problems with the aircraft, which is typically used on short flights and for service to airports with noise restrictions.(Airline Ends Use of Turboprops After 3rd Crash Landing, italics added). A suspicion of possible "broader problems" was poo pooed by interested parties.
Perhaps the NTSB will pick up on that since the icing issue is not yet a slam dunk?
If one landing gear failed to come down, that would put more resistance on one side of the aircraft. Or if one or more would not come back up when they tried to increase speed that would be an impairment.
Even in non-severe icing and just normal icing the plane is somewhat impaired by the ice, so the combination of gear problems at that time could be more of a factor than it would be at other times.
Add that to the fact of a new pilot to that aircraft type (since he had only flown that type for about a month, since December) it could lead to a situation where only an instant was available for a corrective action.
Reacting to a specific aircraft need in an instant is a function of a lot of experience.
It will be a long slog to unravel all these potentially contributing factors.