Co-oximetry has been asked about in Question 26.3 from the second paper of 2008, and in Question 17.2 from the first paper of 2010. Additionally, Question 19.3 from the first paper of 2009 refers to the "oxygen saturation gap" in regards to the diagnosis of intoxication. Generally, the college like to ask about the reasons for a difference between pulse oximeter readings and co-oximeter readings. In brief, one can say that low co-oximeter readings always represent true hypoxia, and high co-oximeter readings always represent true normoxia, whereas this is not the case for the pulse oximeter. The co-oximeter is not confused by ambient light, absence of pulsatile flow, tricuspid regurgitation, carboxyhaemoglobin or methylene blue dye.
Low co-oximeter readings always represent true hypoxia. The core basic science behind the "oxygen gap" between co-oximeter readings and pulse oximeter readings comes partly from the difference between fractional and functional oxygen saturation.
(i.e, pulse oximeter fails to detect that the saturation is normal)
Now, a little bit about methaemoglobinaemia here. Carboxyhaemoglobin, the bright fire-engine red species, is clearly so alike with oxyhaemoglobin in its absorption spectrum that the pulse oximeter will reliably read it as oxyhaemoglobin. The result is that the pulse oximeter reading is a predictable sum of real and "carboxygenated" haemoglobin, i.e. when your sats are 100% you could have a 70% carboxyhaemoglobin content, and your "real" haemoglbin might only be 30% saturated.
Not so with methaemoglobin. Pulse oximetry measurement of mixed normal haemoglobin and methaemoglobin will usually be depressed. When one's methaemoglobin level is in excess of 35%, the pulse oximeter will usually read 85% (Barker et al, 1989). The pulse oximeter will then continue to read this value, whether the oxygenation deteriorates or improves, i.e. at 60% methaemoglobin concentration and a fractional oxygen saturation of 30% or 100%, it will still give you an SpO2 of 85%. This is why highly respected resources report that "methemoglobinemia typically causes the pulse oximeter to report a [pulse oximeter] saturation of ~82-86% (even if the PaO2 is very high)."
Question 19.3 from the first paper of 2009 refers to the "oxygen saturation gap". Though not a formally accepted term to describe this phenomenon, the "gap" is a well recognised feature of dyshaemoglobinaemia. It develops when the pulse oximeter reads a certain saturation, and the ABG machine or CO-oximeter returns a different reading. The "gap" in saturation readings is vaguely representative of the concentration of the abnormal haemoglobin in the blood.
Pulse oximetry is discussed at lengths in a fine article:
Mendelson, Yitzhak. "Pulse oximetry: theory and applications for noninvasive monitoring." Clinical chemistry 38.9 (1992): 1601-1607.
Co-oximetry is a fine thing indeed. LITFL has a nice summary.
Here is the operations manual for an AVOXimeter 4000.
Barker, Steven J., et al. "Measurement of carboxyhemoglobin and methemoglobin by pulse oximetry: a human volunteer study." Anesthesiology105.5 (2006): 892-897.
Mathews Jr, P. J. "Co-oximetry." Respiratory care clinics of North America 1.1 (1995): 47-68.
Watcha, Mehernoor F., Michael T. Connor, and Anne V. Hing. "Pulse oximetry in methemoglobinemia." American Journal of Diseases of Children 143.7 (1989): 845-847.
Barker, Steven J., Kevin K. Tremper, and John Hyatt. "Effects of methemoglobinemia on pulse oximetry and mixed venous oximetry." The Journal of the American Society of Anesthesiologists 70.1 (1989): 112-117.