Frank - here is your evidence concerning CO poisoning and death sourced to combustion appliances, including stoves (sorry they didn't break out non-EPA from EPA) here
http://www.nfpa.org/assets/files/PDF/Research/Carbon_MonoxideDetectorSpacing.pdf
A Center for Disease Control (CDC) study of unintentional non-fire related carbon monoxide
exposures during the period 2001-2003 indicates that approximately 500 deaths occur per year
and approximately 15,000 individuals per year receive treatment in emergency departments of
hospitals (Vajani and Annest, 2005). The nonfatal exposures were found to occur in homes in
64% of the cases and in public facilities and areas in 21% of the cases. Adults over 65 years of
age accounted for 23% of the fatalities, while accounting for only 12% of the population. CO
exposures were not gender biased, but deaths were highly biased toward males (2.3 times
greater). Both fatal and nonfatal exposures were biased toward the winter months. Carbon
monoxide sources were almost exclusively combustion sources, including furnaces, motor
vehicles, stoves, water heaters, generators, space heaters, and machinery.
More important to the rest of us, the report discusses CO detector location in the home as follows (it is a study, therefore research, not code or regulation, but interesting anyways)
Siting of CSD CO detectors should be in every room containing a combustion device. The
detector should be placed high in the space due to the important role of buoyancy within the
source room. If pre-stratification potential exists due to heat sources high in the space
(e.g. heated pipes, solar heated roofing, etc), the detector should be lowered below the
pre-stratification zone but no lower than nose level. If openings exist between the source room
and the remainder of the building, the detector should be placed at the height of the opening or
above to prevent CO dispersion to other spaces without detection in the source room.
The things Google can find eh?