Sure, the USSR can't hit all of these targets with ICBMs. Luckily for them, there's more than one way to deliver a warhead. Not every target in CONUS requires a dedicated ICBM for a few reasons:
For the most part, you actually do need dedicated ICBMs for most targets in CONUS.
See this post here which explains it quite well:
The poster who keeps referring to Managing Nuclear Operations should look closer at the material.
The idea is that the Russians would launch a preemptive strike hitting US silos, and then force a surrender. There are several issues with this scenario. I will tackle them one at a time.
1) Number of warheads.
The Russians currently have about 1500 strategic warheads in service. The number fluctuates to some degree, due to maintenance cycles and other issues, but 1500 is a good place to start.
The US has 450 MM III silos in service. It is a common misconception that you need 450 warheads to attack this force. You actually need many more. When you develop an attack option, your first task is to determine how sure you need to be of a target's destruction. For something like an early warning radar you might only need a 70% probability that the target will be destroyed.
For a silo, it is probably 90%.
So the Russian planner will look at his weapons and their capabilities, and use a mathematical formula to determine how many warheads he needs to send. You look at the WLS (warhead lethality score) to see if the warhead, given its yield, can even destroy the target, if it goes off on target. You also look at the overall reliability of the weapon, it's probability of pre-launch survival, its ability to penetrate enemy defenses, and many other factors. A simplified version of this formula can be found in Managing Nuclear Operations, Page 380.
It is: DE=PK*PTP*PLS*PRE
Where DE is Damage Expectancy
PK is Probability of killing the target (given CEP, yield, HOB, target hardness)
PTP is the probability of penetrating defenses
PLS is Prelaunch survivability.
PRE is Probability of reliable function.
So lets look at a typical Russian counterforce weapons the SS-18 with its 800Kt warhead.
Pk is 1 because it will destroy the target.
PTP is .9 because it is possible that US ABM defenses get a kill
PLS is 1 because this is a first strike
PRE is .8 based on what we know of Russian rockets. They are pretty reliable.
This gives us a DE of .72, far below the .9 we need to provide. So we add another warhead.
The upshot of all this is that those 450 silos will each need between 2 and 3 warheads (depending on the system) giving us between 900 and 1350 warheads needed. Lets split it, and say we need 2.5 warheads per silo to reach out goal. 1,125 warheads. Leaving the Russians with 375 in reserve.
Should be plenty, right?
Well, lets see...
2) Time needed to respond.
On page 136, MNO gives us a timeline of how long it takes for the NCA to make a decision and for the decision to be carried out. It shows that the US ICBMs will be launching right as the Russian warheads arrive. Does this mean that retaliation is pointless?
No.
There are several factors that affect the arrival time of RV's. The primary concern is that of fratricide. MIRVs are limited to an area where all their warheads must land. This is called the footprint.
The Footprint is different for each weapon but it is an oval and can be 100 km on its long side.
It is likely that you will have several targets nearby in a given ICBMs footprint. If one warhead goes off on its target, it can damage or destroy the other nearby warheads. Even if you wait, you have issues, because following RVs will have to pass through the turbulent hot air from previous explosions. RVs are at the mercy of physics. They are unpowered and guided only by the calculations of gravity and drag. Introducing the turbulent and unpredictable environment of a mushroom cloud will have negative consequences for your accuracy.
Lower accuracy means lower Pk, which means we are adding even MORE warheads.
Those 375 reserve warheads are dropping like sentry gun ammo on LV-426.
Back to timing.
So we now know that the warheads can not arrive all at the same time, making it extremely likely that a significant portion (perhaps as much as 50%) of the US ICBM force has survived and is now on its way to Russia
So now lets pause and look at the situation.
The 200 or so remaining Russian nuclear forces are on submarines and cruise missile aircraft.
The US has about 110 (25% of US ICBMs escaped. Lets give the Russians a break here) warheads on their way to Russia. They will be targeted at Russian command and control systems. The reason that this response is picked is because the best option for the US was to cut the Command and Control links between their subs and the Russian leadership.
The US retains about 128 ready SLBMs (the rest are in port and will require between 2 hours and 7 days to make ready) with about 500 warheads.
In the scenario, the Russian leader now calls the US and asks for surrender? Why?
Actually, no, he doesn't.
3) Depressed trajectory
The US attack option would also order the deployed US SSBNs to fire a large number of their missiles at command and control targets. The SLBMs can use a technique called depressed trajectory to decrease flight time at the expense of range and accuracy. Since the targets in this case will not be silos, the accuracy loss is not a major concern.
The other advantage is the short flight time of these weapons. The Russians will have about 3 minutes from detection to the first detonations on top of their command centers.
It is unlikely the Russian leadership will even have a chance to pick up the phone to ask the US president for surrender, before his communications are disrupted. If he is not killed.
In short, the idea that the Russians could execute a first strike that would be effective in crippling the US nuclear forces and forcing a surrender is not grounded in reality.
Even if we assume that the US does not deploy its SLBMs, the US is still in possession of a major warhead advantage, an intact command and control network, and the ability to launch a DT attack using those SLBMs that the Russians would have little time to react to.
The Russian situation is dire. Their surviving nuclear weapons are in submarines that patrol in protected bastions hear their coasts and are unable to deliver DT type attacks. Their command and control systems have been disrupted by the surviving US ICBMs, and they have not accounted for NATO at all.
At this point, the Russians would be better off calling the US to surrender, rather than to demand it themselves.
Targets within 5000km of Soviet bases can be hit with IRBMs (IRNFT won't be signed until 1987/88). That covers European NATO, the Middle East, and North America as far as the Pacific NW.
Which still means you have the issues of range and limited ICBMs for targets in CONUS and outside of that; Moscow will probably want to hit places like Clark AFB in the Philippines and Diego Garcia in the Pacific, for instance. That they can hit targets in Europe with IRBMs doesn't address the fundamental issue already pointed out, given I've only been talking for the most part about American strategic targets.
MIRVs mean not all 2000 of the warheads targeted at the US in this scenario require a separate missile, especially missile fields and densely-packed targets (there were IIRC about 7000 warheads loaded on the ICBM force in the mid-to-late 80s). That allows you to make a few efficiencies.
Except, as already pointed, MIRVs don't enable such efficiencies in targets like Silos, which is why the Soviets had specifically designed ICBMs for such purposes given Silos are hardened. See the above citation which goes over this nice.
Soviet Long Range Aviation was integral to nuclear planning until the end of the Cold War, especially for countervalue targets that you can wait 5 or 6 hours to hit after the first strike nails the key C3 targets. Even if a Bear is a lumbering target that can be intercepted easier than an SS-18, a few squadrons of those loaded out with Kh-20s still gives you leeway to hit a lot of second-tier targets, particularly if the DEW Line and USAF interceptor bases have been hit in the first wave. The Backfires (M1.5 strategic bomber with a good 7000km range) are an even better bet, and there's a few hundred of them.
Which are also limited in range, numbers and capabilities; Soviet long range aviation was already long on the backseat to ICBMs and SLBMs by the 1980s, they never had a capability like SAC and this also runs into the issue it takes 30 minutes or less for an American ICBM or SLBM to destroy said bases before the bombers can take off.
SLBMs have a pretty impressive range in this era - about 7000km for the R29 deployed on the Delta III boats - and even if kickoff happens quickly in an AA83 scenario and a lot of the sub force is at anchor, that gives contingency options for more targets within range on coastlines.
And their bastion areas were before the GIUK and Kuriles, given NATO supremacy at sea with their attack submarines and ASW forces. Said bastion areas are outside the range of hitting South Africa or Australia, for example.
Given the above, I think it's a perfectly rational calculation for the Sovs to make to allocate a relative handful of their 2,000 ICBMs for isolated strikes on limited numbers of key targets in a range of major neutrals and allies further afield. The strategic logic of prioritising Canberra and Brasilia over Rochester and Las Vegas starts to look pretty good once you've already hit the rest of the US.
Not really, given the U.S. is a nuclear power with global military capability while Australia and Canberra aren't, even before you factor in the strategic difficulties of such targeting. For what it's worth the Aussies have declassified their Cold War files,
which revealed they had no intelligence to suggest the Soviets ever targeted their cities.
The value proposition of sacrificing maybe 40 or 50 ICBMs (again, out of two thousand) is: disrupting NATO-allied/-adjacent response (Seventh Fleet can't berth at Sydney if it no longer exists); neutralisating potential threats from major non-aligned powers; limiting the ability of NATO-friendly states to give aid and comfort to NATO remnant forces (highly-centralised regimes like PRI Mexico or the Brazilian junta will be in a very poor position to do anything if their capital is hit), and; cutting off access to key resources that will allow NATO to survive and rebuild post-war (if you hit Caracas or Lagos, how much government will be left to administer the oilfields?).
Beyond the fact ICBMs can't even reach those targets? What is the distance from Kiev to Brasilia?