Here's a hypothetical example of the flaw in this approach. Someone above suggested Japan as a candidate. Let's say Russian and Japan get into a dispute and let's say the Russians grab Hokkaido. Japan threatens millions of Russians and millions of their own citizens in a retaliatory strike in order to defend Hokkaido?
Bad example. Hokkaido is regarded as a very integral part of Japan and even an invasion limited to it could easily be mistaken as a stepping stone in a invasion of the whole of Japan, which
is an existential threat to the Japanese state and would very much warrant a nuclear response. Plus it rather ignores the prospect of Japan limiting itself to just nuking the invasion fleet, since tactical nukes are a part of Massive Retaliation. Then it's in Russia's ballpark to decide if sacrificing 80 million of it's own citizens is worth taking revenge over a ill-conceived invasion attempt. The deterrence game doesn't necessarily end once the first nukes go off.
I rather think I brought up a better example in the Blunted Sickle thread where PDF was talking about how post-alt-WW2 the Entente will be relying on nukes to hold off the Soviets in Europe. He then added the following:
Outside Europe, they're in the familiar model of colonial small wars, and their participation here will be set by what they can afford (likely to be a bit more than OTL). Similarly, the Soviets can stir up trouble to some extent without risking nuclear retaliation.
To which I replied:
The room for the Soviets to stir up trouble outside of Europe is a lot higher because what happens outside Europe doesn't threaten Britain and France in the same way as what happens inside Europe. I mean, consider if the Soviets invade Iran while remaining quiescent in Europe. There is no way Britain and France's colonial forces could oppose such an invasion without getting smashed flat in a war they weren't designed for. That represents a threat to British and French economic interests in regards to the Iranian oil... but that is the only threat to their interests it represents. It doesn't represent a direct and existential threat to their homelands like a Soviet invasion in Europe does and the economic losses are ultimately annoyances, not anything crippling. Now we come to the same problem the US faced with massive retaliation once the Soviets were able to develop the means to effectively deliver weapons to the US: are the Anglo-French really going to sacrifice Paris and London for the sake of saving Tehran?
Similarly, consider if China goes red like OTL, develops nukes and ICBMs like OTL, and in the 70s or 80s seizes Hong Kong with or without Soviet backing. There is no way the Anglo-French could conventionally challenge the PLA on what is essentially their home turf to actually retake Hong Kong. Are the British honestly going to go nuclear over an old colony, potentially getting themselves killed or severely weakened over essentially nothing more then prestige?
This problem is also going to magnify if the Soviets develop the ability to project some of their conventional land forces power beyond their immediate border countries like they did OTL in the 70s/80s.
You know and I know that there are different types of nukes, and a single tactical warhead is not the same as wiping out a city. But most people either don't know or don't care about that, and any nuclear release will be viewed the same way.
However, many* of the people who actually matter when it comes to formulating nuclear strategy
do understand the difference and they hold rather more nuanced views on the subject. The Russian concept of de-escalation strikes, for example, very much hinges on the fact that foreign decision makers would recognize the difference between a carefully selected and limited nuclear strike in certain situations as opposed to a full-scale, nation-killing bombardment. What the majority of people knows does not really factor into the response of the policymakers. Hell, what the majority of people
want does not necessarily factor in to the response of the policymakers.
*Although this does probably exclude the current President of the United States given the thin level of intelligence and knowledge he's demonstrated.
There do remain two problems with a pure-nuclear defense. The first is when you get to issues that might be important to a country but not to the point of going nuclear, even in a limited matter. The number of issues where going nuclear, regardless of the scale,
vastly outnumber those which do. The other problem is that even on those issues which are of such import, a pure nuclear defense leaves one open to things like salami tactics and the like. Not every threat comes in the form of an invasion...