Thanks for that insight, Dreamland. It might have escaped my notice that the King Air and the TBM are dissimilar, so it was good of you to point that out. Every time I make a post, I depend on alert readers such as yourself to keep me on the straight and narrow, and someone usually comes through like a champ.
The thing is, you may get a slow leak somewhere between the oxygen bottle, usually located somewhere outside the pressure cabin, and the rest of the system. The oxygen then slowly fills the fuselage until it finds something combustible such as grease. This is not a common problem; it needs a leak that's small enough to go unnoticed, but big enough to cause a fire after an extended period when the aircraft is sat parked, so that most operators don't bother to shut off the valve at the oxygen bottle.
To ensure against this (uncommon) fire risk, you need to physically shut off the valve at the bottle. It's not enough to just flip a switch in the cockpit, because the leak can be anywhere between the bottle and the solenoid control valve. Of course you must also remember to open the valve before flight, when there will be no visual cue that it's closed, just looking at it. You will need to twist it to make sure it's open.
As you can see from reading the checklist for the TBM, it has what seems to be a check of the shut-off valve itself, there on the bottle, as part of the preflight, along with checking the oxygen pressure at the bottle, which means checking the amount of oxygen available.
There are many accidents which happen to owner-pilots, people who multi-task, when "operating the airplane" may not be what they bring 100% of their attention to. Then you may see rushed or even absent pre-flights because of time pressure.
It's not just a problem with low skills, what we often take to be the case with non-professional pilots. Some non-professionals are real geniuses who can do many different tasks to a very high level of competence, when flying is not all that difficult. The problem then can be lack of full attention to something simple.
For instance, there have been at least two fatal accidents I am aware of where owner-pilots were able to get airborne in just a few minutes, although their gyro instruments needed more time than that to stabilize. They blazed through the start-up and taxi checks and were wheels-up before they had their attitude instruments working properly, when it takes skill to kill yourself that way.
One guy was someone I had flown with just a few months before, spending quite a bit of money on a type-rating on one of his aircraft hoping to use him as a contact for a change of scene. I was pretty peed-off when I read about how he managed to kill himself in his King Air, but from what I had seen during our check ride, rushing everything, I could see how he must have done that.