Things are statistically irrelevant until you become the minor statistical footnote.
Your point isn't that there's no chance of a high CO level in a tank, it's that the odds are very low. Ok, fair enough...
You can do whatever you like. All I'm stating is there aren't enough facts to support any real concern for CO. There are PLENTY of facts to support just about any other way a diver has died. In light of this I'm going to take a jaded stance when someone tries to tell me that the industry/shop/region/whoever is covering those particular deaths up.
Let me turn this whole argument on its head and ask you both (all) to look at it from a different angle.
Among the counter-top display of skills we offer technical divers during training, we include air-sharing... donate the longhose, exit the cave/wreck/depths in a sort of controlled highly regimented formation, et al all in a bid to manage Out of Air Emergencies. This is a skill that is taught as a fundamental and failure to get it down pat results in student failure... However, I would suggest that in technical diving among properly trained dive teams, a real OOA emergency is rare... statistically zero. In 20 years of technical diving I have seen or experienced one such emergency... and that was an open-water diver who followed us into an overhead and then panicked.
A straw poll of a bunch of senior ITs who had happened to standing around shooting the breeze in the headoffice of a technical training agency one time, turned up a similar dearth of actual real-world events. (Among us, we had more than 13,000 dives on doubles in caves and the total OOA situations that required one buddy to breathe off another's kit as per the scenario taught and practiced regularly was zero.
Yet we still teach it and we still practice it.
CO poisoning may be rare, but its seriousness... the result of it showing its face... and the ease with which the risk can be managed... Well, let me put it like this:
"Are you willing to stop doing S-Drills completely... to never bother practicing an air-sharing exit?
Similarly, and germane to this thread (although a moderator is welcome to break this off into a new one), I have approximately 600 hours shared between two rebreathers that I dive regularly. I always carry an off-board bailout bottle... often two of them. I have had to use off-board bailout once. So statistically speaking, carry bailout is not really necessary is it?
Catch me diving a CCR in anything but confined water, and I'll buy you your own unit!