Brian Dunning, the skeptoid, has a post up arguing that locally produced food is not actually better for the environment. His central claim is that getting your food locally doesn't actually reduce the amount of fuel used in shipping, because systems of direct delivery are less efficient than systems with distribution centers. This sounds like a good argument, but it contains a subtle fallacy. Basically, when he shifts from talking about locally vs. internationally produced to talking about direct delivery vs. distribution centers, he changes the topic. Granted, if you look at two complex distribution systems of the same size, the one with distribution centers is going to be more efficient than the one that relies only on direct delivery. But what about a local system that uses distribution centers compared too a global system that uses distribution centers? Dunning assumes that once a system starts using distribution centers, it no longer counts as local. But his own example is of a local system that switched to distribution centers because it was more efficient. The system didn't start importing beef from New Zealand, but Dunning wants us to believe that because one local farmers group switched from direct delivery to distribution centers that it is more efficient to ship food globally. This just doesn't follow.
There are a lot of problems with locally produced food. Peter Singer has a good argument that supporting agriculture in the developing world is just as important as reducing carbon footprint. He also points out that transporting cargo by ship is much much more efficient than transporting by truck. But broad, almost data free, arguments like Dunning's just don't cut it.
In general, calculations of environmental impact are incredibly complex, and the rhetoric of skepticism really doesn't work here. Skeptics and debunkers always frame debates as cases of reason vs. superstition. The tone is "I'm rational, and you're not." But environmental debates are generally not cases of reason vs superstition, but conflicting ways of calculating costs and assessing risks. If one person uses a model that only takes GDP to be a basic good, while another person tries to include, say, the rights and interests of non-human animals, neither side is being irrational. They are using different frame works with different assumptions about value.