kgoan,
I think it became a canard on May 28, 1998. I recognize that not everyone agrees with me. But certainly it became a canard in late October or early November 2003, as knowledge of the AQ Khan network leaked into the open sources.
First off, I don't think anyone is arguing that Pakistan has a "sophisticated" nuclear capability. They can produce fission devices, with yields in the Hiroshima range, that are probably deliverable by missile and aircraft. That's not sophisticated, when compared to US or Soviet model tactical or strategic weapons in the late Cold War, for example. The point is simply that they are sophisticated enough. Detonation of such a weapon in a city would produce Hiroshima-like casualties, which has a deterrent effect.
Second, I wouldn't argue that Pakistan's capability is indigenous. They clearly sought assistance from a variety of sources, and continue to do so. Can they independently design and create a fusion-based thermo-nuclear weapon? I don't know, but tend to doubt it. The point is that over twenty years, they acquired both the technological knowledge and the nuclear infrastructure to produce enriched uranium and plutonium, to weaponize these materials in at least a first-generation fission format, and to not only stockpile these arms but also transfer that proven technology to other parties.
Third, I am talking about the assumption, apparently still shared by some on this forum, that Pakistan has no nuclear capability "in house" - that the Khan labs and PAEC are simply fronts, and that China provides intact nuclear weapons that are the only source of Pakistan's deterrent. I consider that argument as bankrupt as the assumption that the US controls all of Pakistan's nuclear arsenal. Therefore, I call it a canard - it is, at best, an inaccurate assumption based on disputed information, that should be largely superceded by the evidence of the Khan network.
It's all very well and good to boast about scientific prowess - both Indians and Americans (among others) are justifiably delighted in the innovations created by their scientists, engineers, software designers, etc. But innovation is not the only measure of military effectiveness. Sometimes the ability to imitate and replicate is sufficient for many military purposes. The Germans, in the Second World War, created many new military innovations, but lost the war for other reasons. The Allies also were very active with technological innovation, but this was not a panacea - sonar, for instance, did not "solve" the U-Boat problem. The Japanese, who were only able to innovate in very selective niches (24" torpedos and Oka rocket suicide bombs, to name two), still put up a heck of a fight.
My reading on the international (including Indian) response to the Khan network is that it verifies Pakistan's ability to imitate and replicate basic nuclear weapons technologies - in short, a concession that Pakistan does NOT receive its nuclear weapons direct from China. That is the canard I am referring to - the assumption, for whatever reason, that the Pakistanis are utterly reliant on China for any nuclear weapons they may possess. Pakistan undoubtedly received both nuclear and missile technology from China and other sources. They have mastered this technology sufficiently, and have created a domestic infrastructure to support this technology, so that they can now independently produce nuclear weapons without relying on outside technology. The point is that based on the Khan network analysis, it appears the Pakistanis have mastered it to the point that they are capable of independent production - perhaps replicating external technologies and designs, or perhaps with minor indigenous modifications (that might be one argument for the second series of tests on May 30, 1998).
If you're very focused on relative scientific capability, Pakistan's ability to independently create new technology is an important indicator. If you're focused on relative military capability, and especially on the ability to deter through the stockpiling of nuclear devices and weapons, the innovation is not as important as evidence of the ability to replicate and independently produce. My read on the Khan network is that it demonstrates quite decisively that Pakistan can independently produce nuclear weapons components and the devices themselves.
The argument that China supplied all of Pakistan's nuclear capability (and, perhaps, therefore controlled it) might have been plausible when Ravi Rikhye and others raised it in the late 1980s. It might have been possible in the mid-1990s, despite reports by Seymour Hersh and others. It might have been comforting, if perhaps an act of willful self-delusion, after the 1998 tests. But after the Khan network disclosures, and the international response, anybody who seriously thinks that Pakistan relies on Chinese nuclear weapons, and can't do anything on its own, is on the intellectual equivalent of drugs.
That's what I mean by a canard. While I respect the rights of others to disagree, I think they're simply not looking at the evidence.