For what it's worth, the limit imposed both by US NAVY as well as CE standards (in the European Community) for CO2 is 5000 ppm or 0.5% SEV. This is the value used for scrubber duration testing, the loop gas content reaches it and the absorbent is considered spent.
If that value is dangerous or not is being debated, as susceptibility to CO2 not only depends on the actual CO2 value, but also the mechanics of the gas composition (i.e. raised pO2) as well as the ambient pressure (pCO2), and last but not least, the diver him- or herself (just as susceptibility to nitrogen narcosis or hyperoxia). Unfortunately, research in that area is lacking.
All that aside, CO2 monitoring has been implemented in the past (in the Powerbreath Infinito CCRs of which a small number has been sold for military purposes and an even smaller number has found its way into private hands). CO2 sensing requirements are outlined in the CE rebreather standard EN14143, and the Infinito's electronics met them.
The system has been offered to RB manufacturers, but due to the high price (to offset the high development costs) there have been no takers. Due to the fairly small number of units build by most manufacturers (100 - 150 units/year) the price of a say a Meg or PRISM would almost double. The Infinito was in the $15,000+ segment at a time the $ was actually worth something (the electronics come from the UK, Powerbreath was an Italian company).
The Scottish company Divelife also has developed and tested electronics that include CO2 sensing, their system is suposed to be considerably less expensive when it hits the market.
Some German guy advertised a working CO2 sensor, promising its availability (i.e. offered for sale at 1000) at the German 'Dive' this past January ... alas, he and his CO2 sensor were nowhere to be seen.
None of these systems, by the way, can detect CO2 where it really matters: inside the diver.