> From an engineering standpoint, Titan didn't have any safety features at all.
This isn't quite true. It had multiple redundant ways to drop ballast, for example.
What I would say from what I was able to find out (and with some familiarity of safety engineering processes from work; I make cars) is that it's safety-concept was very spotty. It had solutions to some problems, but also large gaps in the safety concept. Safety was not addressed in holistic fashion.
It's interesting to compare this with solutions found in other subs. For example, Titan had four different ways to drop ballast, but from the list I saw, all of them required manual intervention by a non-incapacitated crew and electronics to be working.
On Cameron's Deepsea Challenger--by another rich guy who funded a vanity dive, and relying on homebrew innovations in material science--ballast was held by corrosible wire that would be corroded by seawater in a set time, so the sub would eventually surface automatically. Ballast drop was also triggerable remotely by an acoustic signal, more reliable than radio. The available info is pretty bad, but Titan may not have had those solutions in place.
I'm very much out of my depth (no pun intended) on naval/submarine engineering, and I'm hoping for someone with better knowledge to extend that comparison somewhere.