Re: Even if it was 90% mortality (which I doubt) the human race would survive...
"I heard some organization in the US (FEMA?) once did a study of the results of an EMP strong enough to knock out all the infrastructure in North America. Their conclusion was apparently 80% mortality after the first year. That's WITHOUT ANY nukes going off."
Yes it was FEMA, in the late 1990s. They documented how fragile the US electrical distribution system was and how long it would take to replace the core transformers at risk (several million dollars apiece, 2-3 year construction time, no spares).
The european grid was much more robust - primarily because of much shorter transmission line runs.
This report was in response to the large solar flare which hit earth in 1989 (it knocked out parts of the Canadian power grid) and made people start wondering seriously what effects a Carrington-level solar flare would have (up to that point people were ridiculed as doomsayers, and that's despite a large 1950s event having quite an impact. The difference was in the intervening 40 years electricity had become utterly critical infrastructure).
The USA (and most other countries) have been taking steps to minimise the effects and damage ever since. The biggest problem is induced large DC offsets in transmission lines causing transformers to saturate and overheat but appropriate rejigging of the way things are connected can avoid that happening.
Because the grids are so heavily interconnected an overload failure in one part of the country can rapidly escalate into cascade failures everywhere. It's possible to design more robustness in, but that costs money and accountants won't approve it until they're personally affected.