The diagram in the still of that video is apparently of a cross section of the joint between a machined titanium ring at each end of the main body of the sub, which was wound carbon fiber tube. From the video, the kind of hemispherical ends then bolted to the titanium ring. So the main body was basically just a simple carbon fiber tube.
I've worked with carbon fiber on a very small scale before, and it adds great strength in tension. Depending on the fiber orientation, the strength calculations are pretty simple. In my simple engineering background, I never ran into suggestions of using carbon fiber or other reinforced composite in compression. I can see the carbon fiber adding strength to the tube for lifting and supporting the sub, but fail to see how it makes it any stronger than the binder in compression at depth. Am I missing something?
The other thing pointed out was the joint. Again, my simple background, I can't see the dimensions of the ring AND the tube changing at depth, and I don't think they'd be the same. The stress/strain curves of metals and composites are not similar in my experience. So that glue or epoxy or whatever would seem more like a gasket than an adhesive. Juan pointed that out in the video as well.
I guess I'm restating the obvious, seems like a questionable design at best.