Salon asks the question in the title and seems to imply there is only one answer. However, this question has at least five possible answers--we can call them hypotheses:
1. It (libertarianism) is not really good and people know this and that's why it hasn't been tried before. Libertarianism is a terrible philosophy that no decent human being would support.
2. It is good, but not enough people realize it's good, and that's why it hasn't been tried before. Most people misunderstand libertarianism and think it is what it really is not.
3. It is good in the long run for most people, and most people are aware of this. However, people are unsure what sacrifices they would have to make in the short run to persuade those in power who oppose libertarianism to leave power.
4. It is good, but most people (wrongly) believe that libertarianism, even if it is a well intended ideal, it is just that--an ideal that can never be reached, utopia.
5. Most people's beliefs are a combination of 2., 3., and 4.
Claiming that answer number 1 is the only possible answer is a logical fallacy. We could apply this question to many events in the past and we would understand its absurdity. Imagine a person who is proposing a change X compared to the status quo. If we used Salon's logic, we would say to that person: If change X is so great, how come no one figured out the greatness of X before? Jonathan Finegold Catalan used the development of modern democracy as an example of X, but we can use many other events. For example, if an inventor came with a research proposal to a funding agency, and the funding agency said: if your invention will be so great, how come no one implemented it before, we would all say that this is absurd. Using that logic all human history would have to happen at the moment of the Big Bang.
I lean towards answer number 5. However, all those interested in answering the question in the title of this post themselves should carefully check each of the above hypotheses against reality. They may also try to come up with their own hypotheses.