Logic and the Body Weight Issue.

Through a number of links my attention has been drawn to the works of Gary Taube and particularly his post on how the science should work. I haven't read in detail but this recalls my own confusion on the standard advice on wieght control.
As I have said before, the calorie in/calorie out argument is obviously incomplete at a very basic level. A calorie is a measure of energy - not weight. Hence there must be significant complexity involved in the advice that is being completely hidden. I can't see to get this message across to people around me; my wife is trained in bio-science and just says that of course there is complexity, but it is not possible to describe the physiology to the layman. Her sister was a personal trainer and seems to have swallowed the party line without analysis (along with an Atkins focus - after all, it worked for her).
They may well be right, but I can't just accept it at face value and it is almost certain that any hidden complexity is also likely to hide mistakes in interpretation.

From the beginning then: the basic physics principle of conservation of mass. Body weight must equal mass in minus mass out. I'll take this as a starting point. So to reduce weight one must decrease solid/liquid/gas intake and/or increase solid/liquid/gas outflow. As I understand it the major complexities relate to the tight dependencies between these - reducing intake automatically reduces output until there is no net change. This is homeostatis and is a good thing, if a little inconvenient for those trying to institute a change.
The input end of this flow is the point where we have the most control and I am not going to even suggest that anyone take direct control over their outflow (except in one specific case that I will get to later).

Many years ago I read the Hacker's Diet which had the following image:

which shows the relative importance of various forms of intake and outflow. At first blush it would seem to indicate that water intake is the most likely point of approach. But the volume of water in a body is relatively static (At least, it makes sense for it to be so, I am happy to be corrected). Additionally, it is the prime medium through which metabolic residues - the 'ash' from 'burning calories' - are removed.
Besides, look at the difference between intake and outflow in each of phase. Obviously significant water and carbon are coming in through food and but not going out in solid form.
Here's where I suggest a means for controlling outflow - get more exercise. There is a significant increase in liquid and gaseous loss of material from the body both during and after exercise. Personally I know that I can lose well over a kilogram during an intensive exercise session  - which of course comes back as soon as I re-hydrate. The take-away message being that rapid changes in weight are almost certain to be fluid related and short term.

I seem to have drifted so back to the original point. Control of outflow is difficult and potentially damaging depending on how it is done (I can't think of any healthy way). Control of gaseous intake would be almost impossible and is likely to have only minimal impact anyway.
Control of liquid intake is easy and possible and has the largest impact on total body weight. However, indications are that it is temporary, unhealthy and potentially counter-productive. The first I have mentioned already and the health issues with dehydration are well documented. With regards to counter-productive - apart from water being the medium for flushing waste products from the body, I understand that lack of regular fluid intake results in the body retaining water - hence leading to a higher base weight. I have only anecdotal evidence of these points but the arguments are reasonable.
This leaves control of food intake as being the most effective means of weight control. And recall that I am talking about the sheer quantity of food, not the calorie content. Having said that, it is almost certain that different types of food have different impacts on the body and how long the material remains.
My personal response to this is to (try to) stay with foods that are high in nutrient/weight ratio. By default this appears to result in a diet high in vegetable matter and meats - dense foods and low in carbohydrates which tend to be fluffy fillers. It is interesting how closely this replicates the current fad diets - albeit for different reasons.
There may be a vast amount of complexity still involved in how bodies process inputs and execrete outputs, but there is much that can be achieved through simple reasoning from some basic principles.