Here is a third possible solution to eating in the absence of hunger, and I think the model is a pretty good one. The full-text is available here . The focus is on pediatric obesity, but I think the work could be extended to adults.
Just as it sounds, the model is about giving children freedom and trust so that they can self-regulate their food intake. There are no restrictions on types of foods. This table contrasts this approach versus traditional plans:
The study admits that more testing of this model is needed, but the theory looks correct to me.
Here's on paragraph I found particularly interesting:
"But can overweight individuals recognize hunger cues and stop eating when satiated? During a controlled intervention of 13 thin and 9 overweight adults, hunger ratings following overfeeding were reduced by 41% among thin individuals but remained unchanged for overweight individuals ( 46 ). In fact, when adults were overfed, hunger reduced significantly in the thin individuals and they reported higher satiety ratings. An explanation with empirical support is that overweight individuals both binge eat and diet/restrict food intake more frequently than thin individuals, which lead to disordered hunger and satiety cues ( 47 ). Restrained eaters overindulge in food after perceiving that they have broken dietary rules or have eaten a forbidden food ( 48 , 49 ). In the study when obese participants (who had earlier reported no impact of hunger or satiety cues on their food intake) were obligated to consciously focus on their hunger and satiety signals, they reported that they were able to detect these signals and modify their food intake ( 50 )."
So even with overfeeding, the overweight still felt hungry! It seems to me that even Paleo wouldn't be the solution here, because it is a primary malfunction of the hunger/satiety mechanisms that appears to be separate from the type of food.