In no particular order,
1) Removal of the bridge pieces is now going more rapidly since 1) all the bodies have been recovered, and 2) the NTSB has taken whatever samples they have deemed necessary.
2) It already known that the collapse was particularly "catastrophic" (i.e., the whole thing fell all at once) as it was built in a time period were a lack of redundancies in the design were common. Bridges prior to that period (sorry, I can't define the period exactly) and since that period all have redundancies so that if a part of the structure fails, the rest of the structure is strong enough to support the load.
3) The media's attention at this point is on the gusset plates which connect the girders together. MnDOT had looked into strengthening them earlier this year by adding additional rivets, but this was not done as concerns were raised that the additional holes would weaken the structure more than the rivets would help.
What I want to know: bridges are all rated on a scale of 0 - 120. This particular one had been most recently at 50, but bridges with far lower ratings are still standing. Obviously the rating is not deterministic, but rather statistical. So, what's the standard deviation for any bridges rating? Is it constant, or does it vary with the rating? How accurate is a rating given that they are based on visual inspection? How well do ratings correlate with reality?