Regaining consciousness is unlikely because
1) after consciousness is lost, blood oxygen saturation keeps dropping further due to elimination of oxygen through the lungs and consuming oxygen by the body,
2) N₂O has additional sedative effect besides causing hypoxia via asphyxiation.
The volume of N₂O inside the bag will slightly decrease over time, because some amount of the gas will be dissolved in blood plasma and accumulated in tissues, but this shouldn't result in substantial loss.
That might help with increasing the N₂O concentration and decreasing the O₂ concentration inside the bag, but the given technique introduces other challenges. You'd need to set an optimal flow rate and ensure the integrity of the whole setup in case of possible uncontrollable body movements. I'd rather prefer non-ventilated setup because of its simplicity, keeping in mind that people actually died with non-ventilated bags filled with N₂O or He.
Yeah, mine is approximately 60 cm x 60 cm (in the fully deflated state), and it seems large enough.
Firstly, oxygen is displaced from the lungs, and this makes the concentration and partial pressure of O₂ significantly reduced there. Low partial pressure of O₂ makes the lungs to work in a somewhat inverse mode - instead of taking oxygen from air inside the lungs' space and attaching it to hemoglobin in the bloodstream, the alveoli take oxygen from hemoglobin and release it into the lungs' space. As a result, blood oxygen saturation drops, and when it becomes insufficient for maintaining consciousness, fainting occurs. Consuming oxygen by the body is also a contributing factor in producing hypoxia, but this process depletes oxygen reserves much slower than elimination of oxygen through the lungs during the first seconds of using the aforementioned techniques of gas asphyxiation.
Fortunately, rapid loss of consciousness is known to occur when the O₂ concentration in the breathed gas mixture is ≤6%, that allows us to add up to ~28% of air to a simple asphyxiant without compromising the effectiveness of gas asphyxiation. Since N₂O has an additional sedative effect, it may successfully induce unconsciousness with slightly higher proportions of air than plain inert gases like N₂.
I presume, that might work with relatively small bags, but pumping relatively large volumes of gas (like 15 liters or more) from one bag to another could be quite challenging, considering that you have to do this procedure while having a bag on the head and holding breath (in order to avoid premature LOC before sealing is done). I thought about this method before, and it seems unnecessarily complicated for me.
Deaths occurred even without purposeful sealing. For example, see this
NSFW link.