Well, The Economist is in fact a reputable source, although perhaps as you suggest science isn't its strongest suit.
As far as the causes of autism, I think the following are some of the credible ideas:
Maybe some cases of autism are caused by some funky autoimmune process, say something roughly similar to Guillain-Barre or PANDAS. (and if vaccines are sometimes one cause, I don't think it's the toxicity of mercury, but it could be something autoimmune)
Some metabolic process like PKU.
And I'd say the hypothesis laid out in this article is another credible idea. More and more things are found to have a bacteria cause. 30 years ago, ulcers were viewed as stress or spicy diet. Now people know about the Helicobacter pylori. There are even studies on whether bacteria play a role in causing build up of plaque in arteries. (The article's theory is actually kind of a hybrid in which bacteria contribute to a shortage of sulphur in the body, and this shortage in turn has a metabolic effect.)