In reading the article, these are some of my thoughts:
1. I'm not really into fasting a dog before working it. Especially not fasting a dog for 24 hours before working it. Plenty of research has shown (in humans, so I think the same science can be applied to learning in animals, considering it's often the same things affected) that attention, learning, and memory, are improved in animals/humans that maintain regular nutrition and whose bodies are routinely energized. A dog or human who is fasting for 24 hours has a body that is highly seeking for glucose energy in the form of nutrition, to feed the brain and cells. I could see how working with a fasted/hungry dog would hinder learning more than help it because the focus is in the wrong place (food, rather than learning), it would affect the dog's attention, energy level, and overall level of homeostasis. And because of the dog's (as carnivore) digestive system, they tend to digest their food a lot faster than humans do, based upon the known literature. It's precisely why I feed regular meals regardless of when I might use food in teaching, or if I am working with a dog in the morning I will feed part of a meal first, so that the dog has some energy and sustenance in the body to allow the dog's brain to function optimally. It's like the saying goes: Feed the brain. Your brain (and body) needs regular nutrition in order to function optimally.
2. I keep envisioning this monkey-like dance when picturing the particular hand positions (palm-up, right hand slightly above left, try moving around in this position, elbows out!). Not really relevant, but when I did it in the mirror I couldn't stop laughing.
3. It is just a food-based exercise that teaches the dog to push against you. It's really no different than teaching any other behaviours, except you are shaping a pushing behaviour. Unless you actually pair it with a recall (the behaviour of coming towards you), I can't see it being any more effective than a normal recall that I would teach, and it would take similar amounts of time. I'm actually not that into teaching a dog that it should immediately push against people when the dog is called, I can see that setting up for annoyances in the future. And I don't think I'd want my dog's response to a "come" to come up to me and expect to push against me. I would much just prefer the dog comes to me period, and maybe sits.
I can see a bit of classical conditioning going on as a bond-builder, but really it's no different than any other classical conditioning that pairs human with "good stuff".
The dog is just learning to push against you as a measure of getting food, which is simple R+. It's not what I would do, nor would care to teach. I see it basically as a glorified lure-reward method. The dog sees reward, does the behaviour, and gets the reward. I don't see much teasing going on, nor do I see much competition, as the human has the food, and the dog is working to get it. I don't really like teasing either, but I don't see any teasing going on here, as the dog always gets the food. Teasing implies the dog doesn't get the food or that you are continually showing/removing the food from the dog's grasp, and that's not the case here.
In terms of side effects, it would likely depend on the way it was implemented I think. People use the oppositional reflex for a lot of things - teaching stays, teaching stacking, by applying very slight pressure to the dog in some manner, and then releasing, in a R- procedure. I don't see this as much different. By the sounds of it, it starts so gradual that with most dogs there would likely be little to no anxiety produced. But then again I'm looking at it from a shaping standpoint in which it's done in teensy tinesy steps so that there would be no anxiety induced. I can see the possibility for inducing anxiety if the dog thought that the human was acting as a barrier rather than the dog simply learning to push as a behaviour, or if the dog was teased with the food rather than playing with it and sniffing at it until the food was released. Then there are the dogs that are the exception to the rule, who would not be comfortable doing this type of thing.
It's certainly not something I would care to do, simply because I think teaching a recall on its own is simpler, without the extra "effects", and I don't want to get into pushing with my dog.
But to be honest I'm not really sure it even activates a dog's "prey drive" at all. When I'm using food I don't activate my dog's "prey drive", nor is it induced when I place a bowl of food down in front of them. It is motion and predatory actions that stimulates prey drive, like moving cars, and tugging and shaking, and small critters - things that move or involve predatory action patterns. I can't recall the last time I saw wild animals chest-bumping their prey and pushing them with their chests as part of the motor action pattern.