Windwalker,
First, good on you to be thinking of such an idea. Please do not take what I'm about to say as a distractor for continuing to think of ideas for RB diving.
Ok, let's do an example of a RB dive when things go bad. Diver is cruising along with a setpoint (SP) of 1.2 and everything is fine, and at a steady state. For the purpose of this example let's stay at a constant depth. At this point the oximeter is reporting everything is ok. Now something goes wrong and O2 is not being feed into the loop (several things could cause this). With each breath the Loop O2 content is going to be reduced, but the O2 the diver is consuming will remain the same as long as sufficient supply exists (loop content). So with time the loop partial pressure of O2 (PPO2) will dimish down to .21 and the body will not know anything is wrong. At .21 PPO2 is where trouble begins. With each breath ~.04 (in PPO2) will be reduced. At roughly .1 PPO2 most likely the diver is unconscious and beyond self-help. So that leaves about 3 breathes from the time everything is ok, to too-late. The body of course lags a bit behind, but not enough to provide a significant amount of time to self-help.
The numbers above are generalized as each diver is unique in both consumption, and at which point they would become unconscious. Plus there are other factors, depth, change in depth, temperature that will affect this to a degree. But in general this should outline the problem and thus the requirement that low O2 be discovered quickly and resolved quickly.
But continue to think of how we might solve this issue. High CO2 and low O2 appear to be the biggest causes leading to RB deaths.