>DEPLETED of CO2 => O2 level too high => imbalance, which tends to increase (& become dangerous !)
With respect, this isn't strictly true. If the level of CO2 in the blood decreases (as it does with hyperventilation), there is no concommitant rise in O2 levels. Although it is true that the levels are imbalanced.
The drive to breathe is determined wholly by blood pH and is why some drugs can alter our depth and rhythm of ventilation. As CO2 levels rise, the blood becomes more acid (lower pH) and respiratory drive is enhanced. Of course, vice versa is true also.
The normal rhythmic respiratory effort can and is 'modified' by psychogenic factors - and this is why hyperventilation can occur (eg. in anxiogenic situations) even if blood chemistry is normal. Depletion of O2 in the blood does not drive respiration. For this reason, it is very dangerous to hyperventilate in order to drive out CO2 before a deep dive under water. It is possible to 'run out' of O2 before levels of CO2 (pH remember) have risen enough to trigger respiratory effort. In this case, you'd pass out and drown before even realising you NEEDED to breathe. The blood pH effect can be felt as that dizziness after hyperventilating; the blood chemistry is altering the excitability of nerves in the brain and the tingling is this effect at the periphery.
The reason that hyperventilating is not particularly dangerous is that system is self-regulating. Hyperventilate enough and the change in blood chemistry will cause you to pass out. The unconcious person would then make very shallow, slow breaths (or sometimes not at all) for what can seem like ages whilst the pH returns to normality (CO2 rises to normal levels).
Of course, to the pilot hyperventilation can be dangerous if they were to pass-out whilst in the air!.
Hope this helps!
Paul
BSc(hons) Physiology & Pharmacology
PhD Neurophysiology