Then tell me: how is the exit bag testable, when there is absolutely no indication on how tight or loose should be the seal around the neck? There is just a very vague indication, because it's something that cannot be specified.
And how about the tube from the gas cylinder? Do you know that it will leave a gap between the neck and the bag? How small should that gap be, in order to avoid for oxygen to enter?
I have read of people surviving it and having brain injury; there was a youtube video of such a case. Here I have read of people instinctively removing the bag from the head, after losing consciousness. Most of all, a method that has such poor predictability can cause much anxiety in a situation that is already quite anxious.
How can you say that "proper positioning would remove this concern"? You cannot know what happens after you become unconscious; you might have small convulsions, causing your body to fall on the side. There has been opinions that this method is more suitable if someone is assisting.
Have you read the threads on users that experimented successfully with a full-face scuba mask?
Have you read about the Re-breather and the CO method with formic acid and sulfuric acid?
All of these solutions have a similar principle in common; they can also be implemented with methods very similar one to another.
Cheap full-face scuba masks use the easy principle of "pressure of water on a membrane" to implement one-way valves. I have taken my idea from this.
I believe that testing the flexible plastic tube (either with a one-way valve or with 1 extremity sinked in water) is extremely easy (it's pure physics, or you just need to breath in and out the tube). It's much easier than experimenting with the tightness of the lace that holds the bag on the neck.
The principle of the exit bag is rather easy, but poorly implemented (and that's why people have been looking to the Re-breather and the masks): create a sealed chamber around your head, one tube brings gas into the bag, another tube/opening removes the gas from the bag (which should not allow, in the best case, gas flow in the opposite direction).