One of the other problems with intelligence analysis, is that decision makers frequently do not want to hear what is actually going on - they want to hear that things are going the way they think they should be. This major flaw in intelligence gathering - where agents disregard, or even deliberately misconstrue information that they know their leadership does not want to hear, was one of the major fail points of the KGB. The Sword and the Shield: The Mitrokhin Archive and the Secret History of the KGB is an excellent read, which gives a good picture of how soviet paranoia affected their intelligence analysis efforts.
Of further interest, is of how Mitrokhin came to defect to Britain, instead of the United States, the blame for which can be laid primarily at the feet of the CIA and their extremely timid policies regarding the acceptance of defectors at the time. Especially in light of their disastrous handling of Yuri Nosenko - a KGB defector who established his bonafides with the CIA, and was tortured and kept in solitary confinement in an unheated cell for three years, as thanks for cluing them into the KGB's complete penetration of the American embassy in Moscow. Nosenko would most likely be dead today, or still in his cage, were it not for the efforts of the FBI. The article I link to really soft-pedals the abuse he suffered. The CIA's embarrassment over this fiasco understandably (but unfortunately) made them rather more risk averse when it came to accepting defectors. So when Mitrokhin approached us, we said 'No.' Thankfully, our allies across the Atlantic were generous enough to share some of the valuable information Mitrokhin gave them; information that helped to convict Robert "Rip Van Spy" Lipka of espionage.
Intelligence gathering and analysis is far more than just finding the right information, and verifying its credibility. It is also vulnerable to human foibles, both from analysts unwilling to draw obvious conclusions from the facts at hand, and from decision makers unwilling to hear unexpected or unwanted news. In the Nosenko case, the FBI saved the CIA from their own terrible judgement. Further cooperation and communication between agencies might help prevent or fix more of these cases.
Your last is an interesting sidelight on the current problem of alliance or unilateral warfare. I'd give the US CIA and NRO the highest marks for technical capabilities, but only moderate marks for analytical results (remember 11Sept01?) which makes their liaison with foreign intel groups productive for both parties. But due to concerns about revealing means and sources this liaison is often circumspect in the extreme. But for sure, the FBI isn't going to clue in the CIA when they get Iraq wrong, just as the FBI didn't under similar circumstances with the USSR. The only hope for intra American synergy is concerning counterintelligence, which was your example.
I was hoping you'd have some insights, Celeste. Glad you took the time.
Unfortunately, Tom, Counterintelligence has long been a serious weakness in American intelligence. This has been especially true in the CIA, where CI's lack of resources, position, and access to resources have not served the Agency well over the years (James "Jesus" Angleton was a personal exception who transcended these institutional limitations... for a while).
CI is far more than just investigation (the FBI's specialty), and is better thought of as "quality control" for intelligence. As such, it is of the utmost importance when working through allied services for information collection.
There is no cheap solution, either, as good CI demands the same skills as good intel: language proficiency and understanding of the culture, other sources within that society (and preferably intelligence agency) who can help one triangulate the motives of those forwarding intelligence, etc.
Which means an Agency who relies on foreign intelligence services for information because it cannot itself do collection well, has also compromised its analysis and made quality control a difficult proposition.