Artificial lifeboat scenarios are a bad way to analyze reality. After all, how can you assume that torture will work in that situation?
Psychology shows the destructive effects of power. Torture in the name of "justice" ramps up the destructive psychological incentives and effects of the Milgram and Stanford Prison experiments a thousand-fold.
You can't assume torture would work in that situation but at that point you have nothing to lose. I'm not saying the scenario is likely, just that if there ever actually was such a situation (time critical, bomb on a timer or equivalent type of situation) then I don't really see what the alternative would be. Pain can certainly make people talk even if it is an unreliable way of getting the truth. But in a situation like this, what was said could be quickly verified. Justice doesn't really have anything to do with it. Torture is for getting information, not justice.
Again, I'm not saying this is a likely scenario but it is a possible one. In MOST real-life scenarios this obviously would not be the case. But then again, waterboarding was only (supposedly) used on two or three people out of hundreds of captives. I am not sure of the details but supposedly at least in one case they got info that stopped a major terrorist attack.
So I don't think it's fair to make a blanket statement that torture doesn't work. I would say that it works, but only under a very specific set of circumstances which are uncommon at best.
Can you trust the assurances of the same government that has repeatedly lied about everything else related to its actions in the name of "national security"?
Trust is too strong of a word but I think this has come from multiple sources. How do you know people were waterboarded at all if you don't trust the government about it?