Just curious, why do they fail away from the contact point? Is it to do with de-lamination once they have been over tightened? (all those seatpost and bars are like that, 2 of them very high up)
And just a point that trek wasn't involved in a crash, it just gave out. As are most of them when commented on, Odd that a track bike appear to have had a head on failure.
I think issues like that are because carbon is harder to spot problems in quality control, where as on metal the weld is easy to spot defects in. What do they do for testing? xray? resonance testing, ultrasound?
I don't have an issue with carbon, I quite like it, it the way it fails I don't like, but then Alu can have sudden failure too.