In looking back, I notice I never actually answered Ryan's final question as to how far a hiker can walk unsupported. Obviously, a precise answer is impossible. But certainly we can make some reasonable estimates.
Instead of 40 miles/day, let's back off to 20 miles/day. That allows for 25 miles on most days, but a few short days in case of bad weather or physical or mental exhaustion. 25 miles at 2.5 miles/hour means 10 hours of walking, which is quite a lot. Anything more than that and I think the body or mind or both are likely to get stressed out, opening the door to injuries and illnesses.
Instead of these ridiculous estimates of 6000, 7000, or even 9000 calories/day (who will be the first to argue for 10,000?), let us assume a more reasonable 5200/day. The military estimates a combat soldier needs 4500, so we are being quite generous, in my opinion. One of the posters arguing for 9000 calories/day concluded with a most apt remark about "your mileage may vary". Indeed it may. If you drive a car with one foot pressing on the accelearator and the other foot pressing on the brake, then your mileage may drop under 1 mpg, while your brakes are destroyed in the process. It is certainly possible to do something analogous with our bodies, but I'm going to assume we're a little smarter than that.
Now assume our hiker normally weighs 165 pounds, with 15% body fat, of which 5% is essential and 10% expendable. Let us further assume our hiker bulks up beforehand to 180 lbs. I think I can speak for the majority of Americans that adding 15 pounds is not too difficult. (Yes, there are some exceptions to this rule. Some people simply can't pack on extra weight.) So our hiker will have more than 30 pounds of fat he can afford to burn off.
Now assume our hiker carries a base pack weight of 12 pounds, plus 3 pounds of water, plus 30 pounds of a mixture of plain rolled oats and non-fat powdered milk, so his beginning pack weight is 45 pounds, which is doable even with most ultralight packs, considering that this initial weight will be dropping rapidly and the average weight will be under 30 lbs. Included in the 12 pound base weight is some light salt (50/50 potassium chloride, sodium chloride) and some vitamin pills.
The oats/milk mixture gives about 1700 calories/pound. This includes at least 50 grams of protein and possibly much more, depending on the percentage of milk, which is adequate protein. The oats/milk mixiture does not contain adequate vitamins, especially vitamin C, and thus must be supplemented. The body fat gives about 3500 calories/pound. If the hiker eats a pound of oats-milk/day and burns a pound of fat/day, then he will be able to go 30 days burning 5200 calories/day, without needing to burn any muscle. At 20 miles/day, our hiker can thus walk about 600 miles unsupported before he needs to burn off muscle. At the end of this 600 miles, our hiker will weigh about 150 pounds and look lean, but hardly unhealthy.
Speaking from personal experience, the weight fluctuations described above are nothing extreme. I recall weighing about 175 pounds playing football in high school but 150 pounds when running track. I weighed over 185 pounds at one point during my thirties. My current weight is about 165 pounds. This is for a man who is 5'11" tall and average build.
The estimate of 5200 calories/day was just pulled from the air, and I think it is high. The reason I think this is my reading about how our ancestors lived and how primitive peoples live today. I think even 4000 calories/day is generous. By walking gracefull, so as to avoid wasting energy, it might well be possible to walk 20 miles/day without burning any more energy than we burn when lying about the house doing nothing. Thus it might well be possible to walk two months or 1200 miles unsupported.
The interest in this discussion is not unsurprising, nor are some of the hysterical responses, which are very reminiscent of the hysterical reactions to the early proponents of lightweight backpacking. "Tarps are never going to work!" or "It's suicide to hike in the mountains without heavy boots!" Remember all that? Those of us who were once heavyweight hikers probably reacted thus ourselves initially when we heard of other people hiking with 10 pound base weights. No one likes to change habits and relearn everything from scratch or be forced to replace one set of expensive gear with a new set. But after we calm down, we realize that the change is for the best.
In my opinion, nutrition is the next great frontier in long-distance lightweight hiking. No, the average hiker does not need 9000 calories/day. And no, hikers do not need to cook their food nor will they much care after the first few days whether the food is cooked or not, or spicy or not (though salty-tasting may be highly desireable). All that matters is that the food be nutritious.
What the above example shows is that most long-distance hikers are probably carrying way more food than they really need. Instead of bringing along olive oil and other fats, why not just carry the fat on our bodies and burn it off as needed? During periodic town stops, eat heavily so as to rebuild the fat reserves. Most long-distance hikers actually end up doing exactly this, though not intentionally and thus probably not optimally. In particular, many hikers tend to rush through the towns. What they should do instead is linger a while, at least one day and perhaps two, so as to pack in as many fattening meals as possible before setting off on the trail again. The time lost during the town stops can be regained by walking longer on the hiking days, due to carrying a much lighter pack and not needing to spend so much time preparing and eating meals.