The idea that nuclear war could literally end humanity is a very useful one for focussing the minds of political leaders in why it's not a great idea to "push the button", but it was probably never true, not even in the 1980s when nuclear arsenals were at their largest.
Disagree.
Fallout is frankly a trivial edge effect of modern high-yield nukes; The reason so many radiation casualties occurred in the Hiroshima and Nagasaki bombings was that those early nukes were small and dirty. If you are close enough to a modern nuke to get seriously irradiated, you are probably already dead from thermal and/or blast effects, unless you are directly downwind of a ground-burst (and ground bursts are reserved for toughened military targets like ICBM silos or SAC headquarters - most of which are in remote places with bugger all population for miles in any direction).
The issue isn't getting seriously irradiated by the bomb. Only pony bombs and neutron bombs (which are pretty much pony bombs) have any severe radiation threat--despite the infamous Davy Crockett. (Yeah, the lethal radiation range at full yield exceeded the flight range. Not an issue--the soldiers were expected to jump in foxholes after launching the weapon. Nobody calls fragmentation grenades bonkers because the danger range exceeds the range they typically can be thrown.)
The danger is the overall fallout. When the arsenals were at their peak most of the northern hemisphere was expected to see 20 sieverts of fallout. That's basically a 100% kill of anyone not behind shielding. And note that that also kills most animals we might eat.
Note, also, that real bombs are far more deadly than test bombs. Test bombs are dropped in places with little of value--not that much burns and it's not the sooty black smoke you get from burning a city. Enough of that sooty black some blown high enough (takes megaton-range warheads--Saddam's oil fires didn't cause more than a very local effect because it didn't go high and thus soon rained out) and you'll get it above the rain--it stays up a long time. Also, the standard Teller-Ulam design requires a heavy case--but it's only requirement be that the material have a lot of inertia. Tsar Bomba was dropped at 50Mt because the test had a lead case. Replace that lead with uranium (doesn't need enrichment) and it would be 100Mt and
far dirtier. When weight is at a premium do you think the jackets will be lead or uranium?? They typically tested without uranium jackets.
Besides, the real killer is grid collapse.
A deliberate effort to use the entirety of the world's nucelar arsenals to eradicate as many humans as possible (an unlikely military scenario; Civilian deaths are 'collateral damage' in the effort to destroy your enemy's military capabilities) could probably reduce human population by 80-90%, but that leaves plenty of people to repopulate the world.
Anything that gives a 90% kill gives far, far more than a 90% kill because those 10% can't keep things running.
There are few viable military targets in the southern hemisphere, and plenty of people who live a very long way from any such target, particularly in Africa and South America.
Yeah, there might be viable populations in the southern hemisphere. It depends on how much air exchange there is and last I knew this was an unknown.
Russia isn't going to use nukes just because they are chucked out of Donetsk, Luhansk, or even Crimea; But nukes could well fly if a Russia, emboldened by defeating Ukraine, and confident that the West will aquiesce to further agressive land-grabs, decides to invade or attack a NATO member state.
By a
rational analysis Russia isn't going to use the bomb if they lose in Ukraine. Is Putin rational, though? An irrational actor might convince themselves that they can get away with the bomb in a local situation because the world isn't going to blow itself up over that.
I do agree a Russian win would be more dangerous.