Factors that push it in the other direction are desertion rates if we're talking domestic military occupation and fighting fitness -- how many police and reservists are actually able and willing to occupy a country with a likely insurgency for an undermined amount of time. Some percentage of that 2.9 mil number are not front line combatants. But also if the occupation is only in a single region then the denominator is much lower or if parts of the population side with the military.

Does the figure being "only" 100-1 instead of 300-1 really detract from the point? Would that whole 0.9% even stand on the same side?