Re: "craft's carbon fiber hull"
oompressive stress is a bit easier to deal with than tensile (ripping apart) stress.. This is one of the principles behind pre-stressing structures.
However, you get certain areas that still have tensile stress on them, and this is where failures happen.
* bending inwards of unsupported surface sections
* telescopic compression [most likely this happened]
* seams for things like hatches and cables.
The design has to be a compromise between weight and bouyancy, where you can drop weights to surface in an emergency.
If I might predict what happened, it was a telescopic compression of the 'people tank', starting at a point of stress that had been cycled too many times going to and from the Titanic wreck.
And it would have been rapid, quite possibly causing a diesel explosion of everything organic that was inside - plastic, paint, people, ...
Probably the window and nose cone were found intact for this reason. But the carbon fiber stuff would have shattered. [In the case of a metal sub hull, it would look a bit like a beer can that you stomped on the top of to flatten it)
A readup on the USS Scorpion and USS Thresher accidents might give a perspective on what happens at crush depth...