Kim Jong-il death: Did US intelligence fail?
Modern intelligence gathering, with its satellites and drones operating at the very cutting edge of technology, can often give decision-makers an extraordinary window on the world.
Think of the amazing images earlier this year of US President Barack Obama and his national security team in the White House, watching live video of the raid on Osama Bin Laden's compound in Pakistan as it unfolded.
But the intelligence gatherers do not know everything, and North Korea is a case in point.
US analysts were still uncertain of Kim Jong-il's death some 48 hours after his demise early on Saturday morning.
"In Kim's Death", ran a headline on the front page of the influential New York Times newspaper, "An Extensive Intelligence Failure".
But was this a failure as such or rather just a demonstration of the limitations surrounding the gathering of intelligence about this secretive and reclusive country?
Veteran US foreign policy analyst Leslie Gelb, a former official in both the state department and the Pentagon, says the delay in finding out about Mr Kim's death did not really matter to US intelligence.
"We would not have expected to know something like that," he says.
"What we would have noticed right away is any troop movements associated with an upheaval in North Korea. That's what counts."
In the event, of course, there were no tell-tale troop redeployments reported.
Bruce Riedel, a former CIA officer agrees.
North Korea, he says, "is an extraordinarily hard intelligence target and has been since its creation".
"They practice rigorous discipline in all communications, deception and denial against overhead surveillance and decisions are made in a very small circle."
So, he adds, "being in the dark about a despot's death hour should not come as a surprise.
"More worrisome would be a failure to see signs of regime collapse or preparations for war," Mr Riedel stresses.
Paul Pillar, who had a 28-year career in US intelligence and who is now a professor at Georgetown University, agrees.
"The term "intelligence failure" tends to get used, extremely broadly, to refer to anything that surprises us or, even more broadly, any sudden bit of news about an event overseas when that news does not come to us from our own governments," he says.
"It is," he argues, "only in that broadest sense that Kim Jong-il's death involves any intelligence failure."
Bad 'tactical' intelligence
Are there any grounds for thinking that Western intelligence agencies should have done better?
Here too, Mr Pillar says, there is no basis for applying the label "failure".
"There always will be much the West will not know about a country as closed and tightly controlled as North Korea, no matter how effectively Western security services are operating."
Mitchell Reiss is a former US diplomat who worked closely with the North Koreans.
He ran a multinational organisation established to provide Pyongyang with power generating reactors. In return North Korea was to abandon its own nuclear programme.
The scheme did not work, but Mr Reiss knows the North Koreans as well as any of his US counterparts.
"There was good 'strategic' intelligence on the fact that Kim was ill, had been for a while, and would likely not survive very long.
"But there was poor 'tactical' intelligence as to when, exactly, he was going to die," Mr Reiss says.
He argues that the US has gained a better understanding of North Korea in recent years from debriefing defectors and so on.
But he adds that "it is still a very difficult intelligence target".
"We can always do better," he admits, "but there are inherent challenges, starting with the fact that the US has no embassy in Pyongyang."
'Wait and see' game
Of course, talk of an "intelligence failure" raises comparison with real setbacks: the failure to predict the collapse of the Soviet Union; the false claims about weapons of mass destruction in Iraq; and, more recently, the surprise provoked by the eruption of the Arab Spring.
The Soviet example illustrates the problem that intelligence gatherers are great counters: they can look at missiles, estimate the output of weapons factories and so on. But the underlying political and social dynamics in a society are much harder to read.
Mr Pillar argues that there is no common thread in these perceived intelligence failures.
"US intelligence was making no specific judgment about uncertain aspects of Iraqi unconventional weapons programs until the Bush administration decided it would launch a war and use WMD as a selling point."
As for the Arab Spring, he says "it appears that the US government was already anticipating some such upheaval".
"But the timing and particular spark that would ignite it are the sorts of things it is impossible for anyone to predict, inside or outside the intelligence services," Mr Pillar adds.
Intelligence also provides few clues to guide US policy over the coming months.
Mr Gelb notes that discussions have been going on recently between the US and North Korean officials.
These have focussed, he says, on some kind of deal involving a freeze on certain nuclear activities in return for humanitarian aid. They were reportedly going well.
The Obama administration has little choice now, he says, but to "wait and see if anything of this potential deal can be resurrected".