I question the idea that you can make a sensible estimate of fuel (food) costs for bicycling using an arbitrary $1/calorie price and an equally random calories/mile efficiency for cycling. There's a huge variability to both values. And that doesn't even begin to address that fact that you have to eat whether you ride or not.

It would be far more revealing, to keep track of both your food expenses and cycling mileage over a suitably long period of time.

Since retiring, I've become a bit anal retentive about finances. So I do track just about every dime that gets spent around here. I also track my riding mileage fairly carefully, having set a few goals for myself. No, I haven't actually done the math down to 3 places past the decimal point, but a cursory glance at food expenses during the first year after retiring, when I wasn't riding more than maybe 30 miles/week and the same column in my spread sheet for the past 6 months, where I've been averaging 65 mile.week shows no real increase that can't be accounted for by increasing food prices.

Now the fly in that ointment is that I've also lost 15 pounds (as hoped) over that same 6 months. So where it really shakes out in terms of cycling miles/$1, I really can't say, but I suspect it's considerably higher than 7.8. The question is intriguing question, but it's not one I'll loose much sleep over.