All I see in the picture is some grotesquely obese guy with a poor fashion choice. And he looks pretty cramped in that Ford Excursion. ;)
The article actually specifically mentions that issue, at the end of the first paragraph, "Obesity has caused more people to buy larger vehicles..." I'm sure most of the increase in (non-commercial) car size is due to enormously fat people not fitting into normal cars (and bottoming out suspension, etc). That's pretty much common sense, and we've already had the "bigger cars use more gas" discussion. This article is trying to hit closer to home with the message, "your big fat ass uses more gas". Even a big, inefficient gas guzzler uses more gas hauling around a load of 300lb passengers than it does passengers 2/3 or half that size. Other wasteful behavior is pretty much irrelevant to this discussion, and doesn't negate the facts that body weight affects gas mileage. No matter how much unnecessary cargo, or sand bags, or whatever else you might have in the car (that can easily be removed if you wanted), your ass is always with you, putting additional load on the car's engine. No matter how stupid it is to go grocery shopping in a lifted H2 on 44s with a chrome winch, and an ATV on the roof rack, it still uses even more fuel if the driver is 200lb overweight.
Sure the article says its only .7% increase overall, but that increase is heavily weighted towards obese people. Skinny drivers help bring up the average MPG, so the impact doesn't seem as great. The increase in fuel consumption due to obesity affects *my* gas-mileage by about 0 percent. My 300-lb friend probably makes way more than .7% more trips to the gas station than I do, since he's carrying the burden (so to speak). That extra billion gallons of waste isn't evenly distributed across the population (though the environmental impact of burning that extra fuel to drag one's fat ass around town affects us all equally).
I'd be curious to see the study break out the numbers by weight class, or maybe have a test group of various people driving the same car (or same kind of car, and do it for several models) to see how their specific weights affect mileage. If they showed that being 50-100lb overweight actually costs you (just guessing) 10% more annually, that would have way more impact.
Interestingly, the bigger cars will show less of a performance decrease from obese drivers, since an extra 200-lb on the driver's body is a much smaller percentage of the total vehicle weight. Adding a 300-lb passenger to an 8000 SUV won't really affect much; its a drop in the bucket. Add a 300-lb passenger to a Toyota Corolla, and you can immediately tell a difference in the way the car accelerates and handles. As an obese driver of an H2 sheds weight, there will hardly be any increase in gas mileage; a 100-lb weight loss is barely more than a 1% change in total vehicle weight; mileage might increase from 11 mpg to 11.1 mpg. Drop 100-lb from your body and an econobox getting 35+mpg will gain a couple of mpg. The fact that obese people don't fit in efficient cars dilutes the real problem even more.