Calvin,
It may be the latter - or at least, I may have a bias.
But...
Ring magnets, as ramana has pointed out, are part of the HEU process. And, clearly, they were transferred in the mid-1990s.
The Dutch put Khan on trial for espionage. It may have been some sort of counter-espionage mission, but it seems pretty clear - given recent admissions by the Dutch government - that at least the Dutch feel that the technology Khan stole was functional, and would do what the Pakistanis wanted it to do, which was to enrich uranium.
I suppose the US government could also have been complicit in helping the Pakistanis to hide the fact that their nuclear technology didn't work. It may be that the various revelations in the 1980s, and the Pressler amendment, and the gradual revelations that Pakistan was claiming - admittedly, not PROVING - a capability to first enrich uranium to 10%, and then later 90%, wer all part of a deception scheme.
But you've already got three major states involved in covering up a possibly non-functional Pakistani capability. That's getting a bit thin, in my opinion.
Furthermore, you have Indian intelligence estimates (declassified for the Kargil Review Committee Report, among others) regarding Pakistan's emerging capabilities. You also have India's actions, which do not suggest (in my interpretation) either blithe disregard of Pakistan's capabilities or secure knowledge that they are inoperable.
Iran's centrifuges, according to at least one IAEA report, have traces of HEU that is linked with Pakistan's reactors.
The DPRK, which had functioning nuclear capabilities, sought Pakistan's HEU line.
Iran, which clearly sought HEU capabilities, obtained Pakistan's HEU line, and apparently successfully enriched uranium with it.
Those are some of many data points. They suggest to me, at least, that Pakistan's HEU production is credible - maybe not astonishingly efficient, maybe based on significant amounts of imported technology and materials, but still credible.
On the other side, I see very little evidence to suggest that the HEU facilities DON'T work. It's harder to prove a negative, admittedly. Still, the argument (as far as I can tell, and I'm open to new information) rests on assumptions about a failed test in the May 15-18 1998 period, about which there are substantial doubts that Johann summed up earlier, some trips to China by Pakistani nuclear officials later that month, and a set of assumptions about Pakistani scientific capabilities.
I don't find the latter argument very convincing. If North Korea, which has very limited contact with modern science, can somehow build a nuclear program, I'm not sure why Pakistan can't. We know for a fact that other states, with imported technology, have created missile and nuclear capabilities - Iraq, Iran, Norh Korea, South Korea, South Africa, even Israel. Estimates of the cost of the Pakistani nuclear program in public sources currently run in the $5 billion - $10 billion range, so even resources (over a thirty year period) do not suggest a meaningful constraint (Iraq spent that much in less than a decade in the 1980s).
In short, my interpretation of the data is that it seems to support the idea that Pakistan spent a lot of money on a program, and got something significant for it.
It may indicate hidden (or not so hidden) bias. But I find the "nuke nude" hypothesis more compelling and probable than the "Pakistan bought a dud" hypothesis.
Again, if there are more significant explanations for the "Pakistan bought a dud" hypothesis, I'll be glad to look them over. I've been known to change my mind on occasion
Tim