Just looking at your sourced survey, I conclude the following based on actual population data, and what data can be gleaned from the report:
1. In 1944, the overall population of the US was 138M people, represented by 300 civilians in the survey
2. In 1944, the military population of the US was 12.5M people, represented by 850 officers and enlisted in the survey.
3. Per statistical norms, the civilian pop [1] should be represented by a minimum of 1,068 sample pop. for min statistical accuracy at a Margin of Error [moe] of ±3.0%
4.Per statistical norms, the military [2] should be represented by a minimum of 1,067 sample pop. for min statistical accuracy at an moe of ±3.0%
Therefore:
5. The sample pop's used were insufficient for both sub-groups [civilian and military] to achieve minimum statistical accuracy
6. The survey offers no data on the moe actually used - a major flaw resulting in apparent failure to achieve minimum statistical accuracy.
7. I see no actual summary of statistical data at all, just a verbal report.
Conclusion: An unsatisfactory report. Somebody needed a statistician among the committee members, but apparently thought it unnecessary. Therefore, I consider the report a dud, and should have been strapped to either "Fat Man" or "Little Boy" before it was issued