I think the editor's response is disingenuous. "It's such a tiny number of recipes." Uh, no — they analyzed 18 recipes, and found that in 14 of them the calories had gone up. That's 78% of the recipes analyzed, which is not by any means a "tiny number". (If she'd argued that they used a non-representative sample of recipes, that would be different, but that's not what she said.)