Estimates of the number of people killed in the last century vary from 49 million to 366 million, for a world population that varied from 1.6 billion to 5.9 billion (as of the mid-1990s) with an average life expectancy of between 41 and 66 years. This translates into an average annual death rate of between 30 and 65 per 100 thousand people.
The maximum murder rate, derived empirically from U.S. data, is expected to be 38. The theoretical maximum is 20,208 (with exponential decline) at an average equal to the maximum expected violent crime rate (1,961).
These facts complicate the derivation of population stress, since they imply that the population either extremely stressed or not very stressed at all. My guess is that the murder rate (corresponding to a stress of 3/38, or 8 percent) expresses a low stress/high energy state that drove phenomenal growth in population and consumption. The war deaths may need to be considered separately (though even 61/1961 is an even lower relative stress: 3 percent).