So the analysis from the scientific community is coming in, even as the cleanup and recovery continues.
Over at the American Geophysical Union site, Dave Petley does a good job of summarizing the landslide event itself from a number of other sources. This slide occurred in two distinct events, with the first landslide block becoming saturated upon intersecting the river, and the second block rotating downward to fill the void of the initial motion.
Local seismograph data shows the two motions:
And an aerial on the slide area delineates the two landslide blocks (or their remnants, at least):
And the USGS has already prepared a brief Open-File report based on a 2013 Lidar survey of the river valley. The most telling element of this report, as depicted in the below imagery, is the number and occurrence of landslide scarps and debris arrayed in the area. The slides are delineated A-D (youngest to oldest) with the 'A' slides from 2006-2011 (the Mar. 22 slide is the crosshatched area) and the older slides being in the hundreds to thousands years old.
Was talking to a local city employee on the job site last night. On March 22, he and his son had just crossed east through Oso on their way to harvest firewood. They were paying for gas in Darrington, a few miles east of the slide. As he swiped his card, the power went out. The service station cashier said, "What kind of card was that?" The slide had just occurred.
Fifteen minutes made the difference.