(Earlier, he had told a US congressional delegation that 84 such weapons were unaccounted for.) Others were doubtful that the Soviet arsenal had such weapons. At the time, the general (who subsequently died in a helicopter crash) was a provincial governor preparing to run for the Russian presidency and was not averse to seeking headlines. I decided not to use the information in the profile of him that I wrote for Fortune.
However, eight years later, Google websites are still quoting what is probably Lebed's misinformation.
Despite endless reports and speculation, concludes Frost, "no one has been able to provide, at least in the open sources, concrete evidence of a single case in which a substantial portion of the fissile materials needed to make a bomb had been, or was close to being, illicitly transferred to terrorists, organised criminals or, indeed, anyone actively seeking them".
So perhaps a terrorist group could find the scientists and funding and make their own bomb? Again, highly improbable, says Frost.
The hurdles would be daunting.
The simplest known bomb project - South Africa's six-nuke arsenal - required $US1billion (1980s purchasing power) and 400 people working from the early 1960s to 1982.
In utter secrecy, the terrorists would have to acquire, inter alia, the technicians; precision-calibrated, computer-guided machine tools; weaponised fissile material (either 25kg of highly enriched uranium or 8kg of plutonium); at least 50kg of high explosives; and a supply of kryton switches (for a bomb's detonating circuitry).
Plutonium has bizarre physical properties - six different crystalline forms as it moves from room temperature to a molten state. The technicians would be in serious danger of radiation-induced cancer or, because of its tendency to burst into flames, of being burned alive.
The Rocky Flats weapons core production centre near Denver, Colorado, is known to have had 31 plutonium fires between 1966 and 1969 alone.
Production of a portable device by terrorists struggling with all these hazards is the unlikeliest development of all. The smallest nuclear weapon known to have been built by the US was a plutonium device measuring 26.4cm by 38.5cm and weighing around 22.8kg. Yield was proportionately low: between 0.01 and 0.25 kilotons (at most, 140th the force of the Hiroshima bomb).
Miniaturisation of this nature would greatly complicate development. South Africa's bare- bones nukes identified by the IISS paper as a possible template for a non-state terrorist group's nuclear effort weighed about one metric ton and had a diameter of nearly 65cm and length of about 1.8m.
Frost feels that though the chance of such small terrorist groups as Al- Qaeda acquiring true nuclear weapons "is extremely unlikely", the "risks of their using RDDs could be high".
Such relatively minor weapons of mass destruction would be "unlikely to kill anyone immediately, except via the direct effects of the conventional explosives involved".
But fear of a slow death from cancer could quickly demoralise, and possibly empty, a city.
In analyst Frost's view, however, and despite terrorist Barot's plot, there aren't likely to be too many backyard bomb makers.
Frosts's paper concludes: "Terrorists in general probably share the same ignorance and fear of WMD prevalent in the broader population, and probably see little reason to turn to unknown, possibly unpredictable and certainly dangerous substances or so, at least, we hope."
An optimistic note?
I'd be more comfortable if his optimism were a little less guarded.
Anthony Paul is a former editor-at-large Asia/Pacific for Fortune magazine, now based in Brisbane.