Apparently, Douglas Feith wanted to bomb targets in South America, or perhaps South-East Asia immediately after the 11th September 2001 raids "because it would surprise the terrorists". Following the link you can find much snark as to alternative options that would achieve the aim of surprising the terrorists. It seems intuitively bizarre that anyone would think of such a thing, although there could perhaps have been basis to it.
Why? Well, taking an enemy by surprise always seems like a good thing. The problem is, though, that it's not always that useful. Surely, bombing northern Argentina would have astonished Osama bin Laden. Would it have helped nail him? No. The answer lies in what are called the levels of analysis. To understand wars, four of these are used. The tactical level deals with the direct, immediate confrontation of small units of soldiers, and scales up to formations of them - these days, up to brigade strength. At the tactical level, we are talking about fighting. In the second world war, from a British perspective, this covers everything from Private Snodgrass advancing straight to his front, in Wavell's words, and also in Wavell's words, that he should move like a mixture of a gangster, a catburglar and a poacher, up to the 7th Armoured Brigade commander in his tank turret.
Next is the operational level. This covers the manoeuvres of bigger forces over bigger areas, but it's still about trying to beat the other bastards in a direct struggle. At this level, we are talking about battles. It's sometimes called grand tactics, but mostly the same terms are used that the Germans of von Moltke's staff introduces - because they conceptualised the thing. Using the same Brit-o-scope, the 8th Army HQ handles it, or Slim's command at Imphal.
Higher is the strategic level. This is about trying to win in whole theatres of war, with consequences of national scale. We are talking campaigns of many battles here. Montgomery and Admiral Ramsay putting together the Normandy invasion. And above this is the fourth, the grand strategic level. This is where the political leadership, with the top general staff, decide whether there should be a war at all and who with. Churchill and Alanbrooke negotiating with the Yanks and Stalin.
Now, how much does surprise matter? Well, at the tactical level it is decisive. The reason being that the effect on individuals matters most there. If only a few people, or a few dozen people, are scared, confused and disorientated enough at the right moment, that might be enough. At the operational level, surprise counts for a lot still, but the people who need surprising are the opposing generals. It's not enough to get over the front line - the deep battle will count. At the strategic level, it matters for something (because if the tactics really go wrong, the strategy is irrelevant), but not that much - there is a lot more fighting after the surprise.
And at the grand-strategic level, it's barely meaningful. You ask Hitler, or Hirohito. Surprising the Soviet Union - which the Germans should never have been able to do given the clues the KGB had - was hugely useful tactically and operationally, and at Kiev strategically. But it couldn't finish the job.
So what does this all mean? In so far as the idea of bombing the tri-border area between Argentina, Brazil and Paraguay meant anything, it was to deal with dodgy businesses in this debatable land. But the point of these zones, much loved by terrorists and criminals as they are, is not to locate anything physical there - it's to evade clear jurisdiction. It's not camps, it's addresses on pieces of paper and in WHOIS databases. It would have been a strategic surprise to al-Qa'ida, in so far as they had money or interests there - but it would also have been disconnected from any other activity of theirs. And strategic surprise is of little value - they would have run away, as they did from the botched intervention in Afghanistan. That the enemy came for them there would be uninteresting. The enemy is coming anyway, speaking grand-strategically, and to a lesser degree, the more the enemy goes there the less they go elsewhere. Strategically, other fronts will benefit, and on the operational and tactical levels, well, terrorists and revolutionaries' profession is readiness to flee.
(In a sidelight, the Pentagon seems to have a weird mixture of clue and utter stupidity on terrorism. At once, they realise the reality of open-source warfare, the exchange of ideas and symbiosis between rebels and the illicit economy, but are also convinced that all nonstate actors are really the tools of some demon-state and that there is little difference between groups.)