You can't compare parachutes to diving. No one can fly without a parachute, there is no CEFD (controlled emergency flying decent) but as a survival item one parachute is enough. All military air crews rely on one parachute for survival. If you are jumping for fun then two is a good idea.
Well, I will give you points for deftly avoiding the helmet, seatbelt, airbag examples.
But I have to point out the fundamental flaw to your example of military air crews. This is a major issue and THE reason we even have this discussion. For years, my jobn involved life safety emergency response operations... basicially we run the show when the ***** hits the fan.
Procedural analysis, and incident "hot wash ups" teach a very important aspect of emergency prepardness... risk frequency.
The rarest of events would validate ZERO time, cash, or personal attention if the consequences were minimal, but if the frequency of the event is high even a moderate impact can validate preparation/mitigation.
High value loss events however don't suffer the same sliding scale of cost vs benefit. Lethal results warrant extensive preparation. It is in these instances where a 1:10,000 or even a 1:1,000,000 event can warrant mitigation steps which can seem "overly costly" for such a rare event.
Small businesses rarely survive the downtime from fire damage, and as a result sprinkler systems, smoke and flame detectors, alarm monitoring, or even flame-retardant/resistant materials can be justified to a fairly high level. Inventory losses from internal theft will rarely cripple the company unless it impacts production or sales. As a result of this, security is not warranted by most small businesses as the cost is too prohibitive.
When it comes to the loss of life, everyone will/can put a different value on this loss, and by direct relation, the amount they're willing to spend in prevention.
The reason a military air unit may use 1 chute instead of 2 is not that a second chute is pointless, but instead, their risk analysis shows its less costly than just paying the survivor benefits. The fact that most military pilots will NEVER use a chute/ejection seat makes the low incident factor suitable for 1 chute use. Keep in mind however that they will ensure to use the SAFEST chute design (no airfoils) and open them on automatic systems to remove human error. Jumpers on the other hand will carry a second chute. There may be examples of single chute jumpers (say low altitude, static-line jumps) where reliability is high, and even in the event of a failure the time involved would be too low to deploy a second chute anyway.
I'm of the personal opinion however than unlike insurance companies, military policy makers, or injury lawyers... I don't believe a value can be placed on MY life.
Even a rare fatal incident seems a little to frequent for my taste. As a result, I'll gladly carry gear that can reduce the event of incidents occuring, as well as those that allow me to mitigate the impact such possible events may have.
Carrying a pony bottle is not a significant cost for the benefit it can provide. Would I use one at 20' of seawater? yep. 20' of pool water? Probably not. Why? because pools have few entanglements, visibliity issues, sharp rusted metal, or random sea/plant life to deal with.
Outside a controlled environment the pony bottle or even twins are easily justified.
Just because "everyone else" does something for years with 'minimal' fatalities, doesn't mean doing so prudent. Fatalities happen because of a series of breakdowns occurs in series (which applies to a lot more than diving BTW
). Mitigation of risks can help break that chain of death and save a life.