You can read what Jenny Ruhl has to say about water loss and gain here:
http://www.phlaunt.com/diabetes/43067073.php
Ah, so..
Whether you are low carbing or not, you must burn off 3,500 more calories than you take in to lose a pound of fat and you must eat 3,500 calories more than you need to gain a pound.
A rather gross oversimplification. Not all calories are created equal, and we don't process them the same way.
If the body can't get glucose from the diet it has two choices: use stored carbohydrate--our friend glycogen again, or convert dietary or muscle protein into carbohydrate using a lengthy process called "gluconeogenesis" which takes place in the liver.
..and elsewhere. Like kidneys, via lactates and the Cori cycle, which is why metformin's contraindicated for people with impaired kidney function as it can lead to lactic acidosis. But unless we're starving, the body doesn't want to cannibalise itself and burn muscles. So glucose levels drop, pancreas secretes glucagon to signal fat cells to give up triglycerides, and then gluconeogenesis can convert that into glucose. This is the bit that's impaired in us, and why LCHF/Ketogenesis works.. Burn fat, not carbs!
Strange that Ruhl overlooks preferential fat burning to muscle given ample evidence from fasting experiments, voluntary or enforced. And from an evolutionary point of view, it makes zero sense to suggest we would consume muscle in preference to stored fat.
If you eat a significant amount of carbs, your liver and muscles grab glucose from your bloodstream to replenish that emergency stock. As they do this, four grams of water join each gram of glycogen and, as fast as you can say, "Omigawd, I cheated!"
I think someone has.. cheated. That only happens if liver and muscle glycogen stores are depleted. As muscles don't like giving it up, that means exercise, and as glycogenesis requires energy, well, that's another benefit. But the water isn't gained or lost, just redistributed.. Otherwise athletes would keel over from hypovolemia as their muscles dehydrate glycogen. The metabolism really doesn't work that way.