Personally I'll believe almost any scenario that does NOT require the airplane to have gone "straight down" when the problem occurred. I believe this lets out flat spin, and may even let out normal spins and spiral dives.
Why? Because if we believe the ocean current drift data and see where stuff was recovered on the 6th and back-plot it, it ends up about 30nm before the 02:10 ACARS fix. So if the problem happened instantly at 0210, the plane has to turn around in zero distance and cover 30 miles backward. Since it isn't likely it turned around in zero distance and it isn't likely the problem happened exactly at 02:10, the plane probably covered closer to 100+miles horizontally from failure to impact.
Now, let's come up with a failure scenario that starts at 02:13-02:14 or later and ends up 30nm south of the 02:10 fix.