Slate's story focuses on the question of the cost of labels. If it's just ink, the cost is virtually zero. However, if labeling changes consumer behavior (or retailer behavior), costs may change substantially. The story, perplexingly, doesn't discuss the question of who will absorb those costs--if non-GMO food costs more to produce, presumably at least some of that will be passed on to consumers, who will subsequently choose whether to buy the cheaper GMO product or the more expensive non-GMO product, unless producers give up on GMO products entirely. To the extent that consumers respond to price changes by decreasing their demand for non-GMO foods, the magnitude of the costs of changing production to meet the new demand will be lower.
I covered a similar issue of labeling-induced production changes in an article on disclosures and the meaning of "falsity."