There's a significant difference between Fat Man (21 kilotonnes TNT yield) and Little Boy (15 kT) and 400 Minuteman IIIs each carrying a W87 (300 kilotonnes each, 120 000 kilotonnes total). And that's one third of the US's current nuclear triad. The second leg is 14 submarines, each carrying 20 Trident II missiles, each armed with an average of four warheads in MIRVs, for a total of some 1100 deployed independently retargetable warheads, which can be 475-kilotonne W88 warheads or something smaller. I'd look up the gravity-bomb numbers, but I think I've had enough already.
The effects on a country of 1000 or 2000 radioactive mushroom clouds seem like they'd be quite a bit larger than the one or two we've seen previously. You could see a billion people dead within an hour.
But they're very unselective weapons. The reason for the trend toward these sub-megatonne warheads is that it makes them more selective so they have more strategic value. But they really can't compete with simple precision weapons there. The US's force of 2800-some warheads costs about US$60 billion per year to maintain, about 20 million dollars per year per warhead. US$200 million will buy you 200 000 commercial drones with which you can kill almost 200 000 individually selected people with grenades. That's enormously more strategically valuable than the million random people you can kill with the warhead.
You need people to arm and deploy 200k drones. With nukes you just press a button. They're also a lot cheaper if you skimp on maintenance and make up for it in quantity.
What? Nobody has been nuked since the first atomic bombs in Japan, and those were devastating, convincing them to fully surrender. A lot more powerful and effective ones have been developed during the cold war. A couple megatons dropped on a major city could cause a massive firestorm. EMPs could be generated high up in the atmosphere knocking out the grid and unshielded electronics.
