On the possible death toll percentages.
from the 1980's doom's day book (a nuclear war briefing - now declassified) its based on population density.
the east and west coasts of CONUS would be nearly depopulated. either from trauma, disease, starvation, lack of clean water or evacuated.
-
the carrying capacity on the west coast is a problem in the south due to lack of water. the typical buildings in the south and south east - the new ones - don't have windows you can open. Summers would be hard. the new construction - especially high rise - from PA to ME would be nearly impossible to heat in the winter. In the midwest, the river valleys would be viable - but the upper plains away from the mountains would be hard to survive in during the winters.
a huge number of people wouldn't be able to adapt in time. add to that;
nearly 20 percent of USofA inhabitants have some continuing medical requirements for extending their lives.
10 percent require direct care of some type.
20 percent are children under 15.
20 percent are elderly
there is a lot of overlap in those numbers but vulnerable populations wouldn't survive even a short term outage - based on hurricane Katrina's numbers. After Katrina help arrived to the majority of the affected within a few weeks.
the math is ugly.
EMP is also one of my major worries. low probability, high impact - and i would add, very long recovery time.
One note - i see alot of people, who should know better, talking about a single bomb event - a salted tamper high gamma yield device would turn the lights out LOS (so one bomb over Kansas) but the E1, E2 would require 6 or so typical thermonuclear weapons to fry all the electronics. the grid bombing wouldn't be a problem for Russia or China - but without help DPRK or Iran might not be able to pull that off. one bomb - no lights or water. 6 bombs, no trains, cars, air transport and no electronics at all - radio and computers would help stitch our country back together faster.