I can see that you take the supposed rendezvous in Prague pretty seriously. A few points about this: First, if the reporting about such a meeting was halfway credible, the administration never would have stopped adducing it as proof of the Iraq/al-Qaida connection. There have been so many instances of stretching information to fit the case, they certainly would have kept this one front and center. (See, for example, Colin Powell's misreading of the recent Bin Laden tape. The message really showed the divergence of Iraq's and al-Qaida's interests. Saddam wanted to avoid an invasion; Bin Laden desperately wanted a war to drive home his point about an America campaign to destroy Islam.) There has been so much pressure on the agency to deliver intelligence that shows an al-Qaida/Iraq link that it is inconceivable that this "can of worms," as you call it, could be left unopened.
Contrary to what you suggest, the CIA would not be embarrassed by the fact that a meeting occurred between Mohamed Atta and the Iraqi intelligence agent al-Ani in April 2001. There was no reason for Atta to be on anyone's screen—except perhaps the Germans' since they knew about the Hamburg cell—so this cannot be counted among the various screw-ups of the period. Even today, Langley would be thrilled to confirm this. The Department of Defense's recently created intelligence unit, which second-guesses the CIA right and left (and, I'm betting, pushed the now-notorious forged documents about sales of uranium by Niger to Iraq) would never let this pass if there were any reason to believe the report. And you're giving the agency too much credit for being able to keep a secret. If there were anything to the report about a meeting—or if there were any kind of subterfuge going on—someone at Langley would leak it to the Washington Times or Sy Hersh. The checks and balances on government power take many forms.
You are right to be skeptical about the FBI's ability, especially in years past, to follow the thread of a conspiracy to Baghdad, Peshawar, or elsewhere. But the same is not true for the CIA or for the U.S. Attorney's office in the Southern District of New York, where the staff was extraordinarily talented and dedicated. And while we criticize the bureau for failing to "connect the dots" of the early terrorist conspiracies of the 1990s, one point we make in Sacred Terror is that the problem for the government as a whole was recognizing that a paradigm shift had taken place—that the real terrorist threat came increasingly from "non-state actors," not from states. The counterterrorism community was straining to collect every bit of evidence of state involvement in terrorism—from Iraq, Iran, Sudan, and others—throughout this period. In fact, there were plenty of officials who would have been only too delighted to find real Iraqi complicity in some terrorist plot; it would have made it much easier to shore up international support for sanctions, which were eroding, or military action. The solid connection between Iraq and Bin Laden, however, was not there.
We could bat around the different accounts of the Prague meeting forever. I'm convinced—by my friends in the intelligence world and the accounts not just in the NewYork Times but also in the Los Angeles Times, Newsweek, and elsewhere—that the source of the al-Ani/Atta story recanted and Czech animadversions to the contrary are about tail-covering, which happens on the Vltava as much as the Potomac. (And yes, as you can tell from the State Department's Patterns of Global Terrorism 2000, we were on top of the threats to the Radio Free Europe/Radio Liberty in Prague.)
There is a much more important point here: You write that, "it remains an unfortunate, if tautological, reality in counterintelligence (or in weapons verification) that nothing that is successfully hidden is ever found." Well, sure. That's pretty much the Rumsfeldian "absence of evidence is not evidence of an absence," and, I hope you'll agree, is not a basis—or even, really, a partial basis—for waging a war. This is, after all, not an abstract discussion.
The lack of a solid story on the meeting in Prague, the very incomplete picture of the Iraqi relationship with Ansar al-Islam and its relationship with al-Qaida, the allegations of occasional meetings in the 1990s (which are not news), this is very far from being the kind of substantive relationship that would make really credible the administration's hypothesis that Saddam might give a weapon of mass destruction to al-Qaida. And, it should be added, this is less than an iota of the kind of information that we have on the relationship between other terrorist groups and their state sponsors.
You say that we don't need to go too deep into Saddam's logic to find a motive for him to attack the United States. But if we're going to put American lives on the line, of course we have to figure out if the postulated threat is real. Evidence of the relationship is lacking. The motive still seems implausible to me. Saddam has shown an incredible will to hold on to power. And while he does miscalculate, his goals—Kuwait's oil or the destruction of Iranian power—usually make sense in the context of his regional ambitions. Sure, he has said menacing things, but the cost and benefits of killing lots of Americans in a terrorist attack and risking his regime don't calculate. He knows that hitting us would not cause us to desist from bombing his forces, keeping sanctions on his economy, or trying to subvert his regime. Until 9/11, he was doing just fine by not provoking the international community and letting the sanctions regime fall apart. It's ironic, but Osama's triumph was actually the beginning of Saddam's undoing.