My instant reaction to the title was that it's a preposterous hypothesis, as I'm someone who thinks that groundballs are underrated by the standard tools of internet sabermetrics, especially FIP. Saber types have tried to factor batted ball percentages in with tERA and SIERA but they've got plenty of their own drawbacks. But when I look at the stats of some prominent ground-ballers (Webb, Wang, Lowe), the FIP-ERA difference isn't noticiably large, except in Lowe's case, where he has generally had a much bigger ERA than FIP over his career. This is who James points to as the exception to his hypothesis that GB pitchers are overrated!
The main thrust of the article is that extreme GB pitchers seem to have a much higher risk of injury, which I kind of think of as an entirely different axis from what I think of as "overrated" anyway.