I'm curious if the window remained intact. This seemed like a logical failure point.

My guess would be that the window wasn't the failure point, assuming that pressure hull was indeed found in many pieces. If the window did go first, the pressure delta on the rest of the hull immediately begins to equalize, so the amount of force the hull is under immediately goes down, making failure of the hull instantly less likely. The inertia of all the water rapidly entering the vessel might do some damage, but that inertia would have to overcome both the pressure of the ocean pressing in on the outside of the hull, and the tensile strength of the carbon fiber, which is stronger in tension than compression. So my guess would be that if the window failed, it would result in mostly just the titanium end caps being blown off rather than a complete destruction of the hull.

If the hull itself failed, however, given the way carbon fiber fails, finding the hull in many pieces seems to be the expected result.

The window had worked fine for dozens of past dives - just because something isn't "certified" doesn't mean it won't work reliably.

The carbon fiber hull was a much bigger concern.

Something working in dozens of dives does not imply it keeping to work in the next one.

You can recycle your argument for the hull...what's your point?

Unlike the window, the carbon fiber hull had to be replaced due to stress fatigue. It was a known issue on the sub.

If the window failed the sheer momentum of inrushing water would probably blow the opposite cap out. Or even tear the whole thing to pieces.

The hull is just as likely a source of failure - there's a reason why deep diving subs use spherical metal pressure vessels.