Interesting and deep question.
Before I read the body of your question, I thought: Yes, because doing things like yoga have de-stressing psychological effects that lower cortisol and allow your hormones to function more optimally. Conversely, if you obsess over food and nutrition too much, then it can have the opposite affect, and work against you. Chris Kresser discussed this with Kurt Harris during one of his podcasts.
But you are talking about psychology in a different sense. You're talking about using BF Skinner style conditioning. Before I object, let me first say that I support self experimentation with (almost) anything that anyone finds appealing, so long as trying it for two weeks or so isn't dangerous or expensive despite its potentially large upside.
That said, I'm dubious that this type of conditioning will work for most of us, and for 2 reasons primarily:
Insulin. Just smelling food produces insulin. So, I imagine, you smell the food, insulin starts pumping, the nutrients and sugar in your blood get cleared, and now you have legitimate hunger to deal with. The counterpoint, however, is that if you persist through enough iterations of this experience without caving, then perhaps the insulin response will decrease, and, in the long run, you will become less susceptible to such temptations. But: it doesn't seem practical to go through this unpleasant experience with all of the bad foods that one finds tempting. It seems much easier to, say, eat the type of foods that don't result in large insulin spikes.
Food reward. When we expose ourselves to food-related rewards, which include both taste, texture, color, and smell, repeated exposure will not necessarily result in a dulled response to those reward-related characteristics. Quite the contrary, they can be tagged in our brains as related, and thus hardwired as concomitant characteristics, which means that every time you smell that chocolate cupcake, your brain expects a lot of calories that taste like fat and sugar, and your insulin starts pumping as part of this anticipatory process. Eating those calories without the associated rewards is one way to decouple the tagged association in your mind; this is what Seth Roberts talks about. It's not absurd to think that conditioning might produce similar results. However, the nature of addiction suggests otherwise. If we avoid a temptation numerous times, but give in to it every once in a while, that's all we need to reinforce the tagged association in our mind. Consider the example of checking your email or RSS feed obsessively. Most times there are not new emails. But a few times there are. And this reinforces the time-wasting practice of checking email obsessively. Thus, with conditioning, if you give in even very infrequently, then the association will remain intact. And, if you never give in, then you have already overcome the temptation, but you've done so in what seems to me as the most unpleasant manner imaginable.
Conditioning seems too close to brute will power to me; it's just concentrated experiences of intense will power. Will power concentrate sounds worse than will power proper, to me. What is brilliant and clever about the paleo diet is that it encourages foods that help us overcome these temptations unconsciously or effortlessly. For example, high protein, high fat meals satiate people, which makes them less susceptible to other tempting foods, even in the long run, because their insulin levels remain low, and their fat burning energy pathways are developed and functioning. In my opinion, conditioning seems to run counter to what is so brilliant about the paleo diet. It's a difficult alternative to an easier strategy.
Summing up my opinions here, surely psychology is a huge factor in weight loss, just not in the Skinnerian conditioning sense of the term.
ADDED: The author mentions the requirement of staying on the wagon with paleo and paleo-like diets. However, he fails to mention that lots of tempting foods lose their attraction over time on the paleo diet. For me, e.g., pizza now tastes like cardboard, bread tastes like a boring platform for butter, beer now seems sour & heavy with a simplistic flavor profile compared to wine or whiskey, and most sugary things are now cloying. A week ago, I actually gagged on sweet tea that I thought was unsweet tea.
Also, the article lacks an evolutionary perspective. It pretends like our susceptibility to unhealthy, fattening foods has been constructed arbitrarily, and that we can simply be conditioned to like or dislike various food characteristics just as arbitrarily. The problem with this tabula rasa notion is that it assumes that we didn't evolve to respond to various food characteristics in specific, hard wired ways, which are only plastic to a certain degree, and are based on features of the environments to which our ancestors adapted. And, of course, most of us here on PH would not make this assumption.
I may have read the article too quickly, but I don't see how any of the evidence lends itself to exposing ourselves to food cues without eating. Everything else the author writes is basically in line with what most of us on PH believe. This additional strategy, however, seems unmotivated. How does triggering the expectation for a food, then not getting it, lead to the diminishing of that association? I suspect that the author thinks that this has been explained, but I don't see it. Worse, some of those associations will hold with foods that we do find acceptable. So, for example, the fact that I eat dark chocolate or sweet potato maintains certain characteristic-reward associations that chocolate cake presents.
To sum up my additional comments: the insulin and food reward arguments do not lend themselves to the de-conditioning strategies described. The role of insulin will perform this decoupling naturally by simply making us less susceptible to reward-related temptations, as the author recognizes. And the role of food reward will also result in this decoupling naturally, but not by associating reward characteristics with nothing - which would not obviously result in decoupling - but by eating foods that provide the calories without the reward, which will decouple because our brains will associate those calories with characteristics that are less rewarding or fewer in number.
