It's something Dr. Norman Friedman (one of the world's leading experts on this sort of thing) said when presenting a paper on flexible response at the USAF Academy. Flexible response (dumb idea by the way) was based around the concept that each side would let off nukes on a tit-for-tat basis until somebody cried uncle (said it was a dumb idea). So, Dr. Friedman proposed, we needed a target system that, if adopted, would confer a major military advantage to the United States, would great increase the efficiency of our war effort and enhance the power of our diplomatic presence while giving no option for an equivalent response from the opposition. There was only one solution to that requirement.
(Profound silence)
The answer is to laydown 350 kilotons anywhere inside the Washington Beltway. It achieves all the above but the Soviets cannot psychologically make the only possible response (nuking Moscow).
Dr. Friedman was carried out of the lecture theater shoulder-high by cheering students.
On the subject matter, several nuclear devices that were lost in accidents have exploded. In the nuclear weapons business, "explosion" and "exploded" refer to situations where the conventional explosives surrounding the "pit" (the physics package at the center of the device) go off but they do not start the chain reactions in the pit. The pit going off, resulting in a nuclear fireball is an initiation. (The technical difference is that normal explosives going off is a chemical process in which a small volume of cold explosive becomes a large volume of superheated gas. A pit initiating is a physical change at the atomic level which releases massive mounts of energy. If you like, we can go into that in more detail). the combined process is usually referred to as a detonation or a NUDET for short.
We've had several cases of where the explosion of the conventional layers cause a partial initiation of the core. This is called a "fizzle" and is generally speaking a disappointment. Usually, the yield is three or four tons. The highest non-nuclear fizzle I can remember offhand was about 30 tons. This is way short of the 150 - 350 kilotons that we normally regarded as desirable. Most of the Nork tests have been fizzles; one so much so that there was a lot of doubt whether there was actually an initiation component at all.