Briefly
If on alternate days you cycle and rest, and your ride uses, say, 1200 calories, what's better, to eat 2600 calories daily (and then where does the deficit come from on the riding days?) or to alternate between 3200 and 2000 calories (and what kind of diet supplies you with 3200 without disturbing your digestive system?).
For the impatient, let me illustrate the question with a figure.
Not So Briefly
My cyclocomputer tells me that my ride used 1200 calories. I'm here not trying to determine whether the figure is accurate. I'm trying to better understand the appropriate diet for a workout that uses such a substantial fraction (or, for some, a multiple) of the basic daily caloric needs. The other question (how can we—or how does a cyclocomputer—determine the calories used given distances, altitudes, heart rate samples, and speeds) is a very interesting question, but it is not the subject of the present discussion.
Since going through what was most likely "hitting the wall," I've been cautious to eat more, especially carbs.
During the ride I can get 350 or so calories from three small snacks.
Right after the ride I attempt to use this glycogen replenishment window and consume 170 g (about a third of a pasta package) to get ~550 calories heavy in carbs. With a sauce and fruits this after-the-ride lunch might add up to 800 calories, which is not too far from what an adult might need without a workout (assuming vaguely 400+800+800 for the three meals).
The purpose of this worked-out example is to illustrates the difficulty of attempting to reach 3200 calories in a day.
If I did this route daily, it would be simple enough. Rather than the daily ~2000, I'd aim to consume ~3200 calories daily.
But, in full humility, admiration, and respect for those who can climb three mountain ranges racing one day and still be expected to repeat the feat on the following day, I need a day to rest.
What do you do if you ride a day and rest the next? Is it more sensible to increase your caloric intake to, in this example, 2600 daily, or would you alternate between 3200 and 2000?
I am not trying to lose weight. More importantly, I'm hoping to avoid that the caloric deficit would come from losing muscles.
If you choose option 1 (a regular diet), where does the deficit on the training days come from?
If you choose option 2 (an irregular diet), is the human digestive system able to handle the stress of this fluctuation? Or is the secret really to drop the quality of the food and eat, say, cakes or chocolate bars, and then a thousand extra calories become a small volume of food. In that case it seems a spreadsheet would be necessary—to avoid that the outlandish number of calories do not come with an equally spectacular amount of saturated fats. Few would complain of the opportunity to indulge in extra slices of cakes and chocolate bars. The trouble then is the accompanying (saturated) fat. Even if it's not the objective, one of the nice side effects of a workout is to develop a larger heart muscle and larger-flow arteries. It doesn't help if that happens while both become surrounded with fat deposits.
Update
Part of the problem is fear that my hunger triggers may be entirely off. They only signal a very slight extra hunger on riding days, which may mean a slow, inadvertent, and gradual muscle loss. The combination of potentially being hard to measure (slow) yet cumulative (gradual) muscle loss is what makes it worrisome.
Judging by the very existence of hitting the wall, it's not that hard to exhaust the carbs stored in the body. I understand that these carb stores increase with exercise, but meanwhile what mechanism does the body use to average out? Are fat stores really capable of building up and getting used in a short-scale cycle?