Pollan argues that one way to define food is to ask yourself if your great grandmother would recognize the item as food. On the surface, this is a good way to determine if something is "food". What you miss though is that humans have been special engineering food items much longer than we realize.
Karo corn syrup makes it's first appearance in 1902. Vanilla extract shows up in 1847. The first synthetic baby foods start showing up in 1867, and margarine shows up in 1870. Now granted, it was probably less common for great grandma to have vanilla extract, Karo syrup, and synthetic baby food. If nothing else, the cost of those products would have placed them out of the reach of many people in great grandma's day. So why did these products continue to lead to other products? Time. We forget that key component to the equation.
No microwaves, moderate to limited availability of ready made food sources, in short if you wanted to eat something, you needed to make it. It wasn't just food that took time. Every household chore took longer. Washing, mopping, cleaning out the oven, and food preparation required serious effort and labor. So the evolution of these "easy" products was in direct response to the need to cut time somewhere.
So if great grandma had various enhanced food items available, what does that do to the food rule? Just how far back do you go in search of "real food"? That's going to completely depend on your time limits and your desire to go back. I tend to look at the middle ground of things that both great grandma and I would consider "food". While she may have considered lutefisk a food item, I tend to steer clear of foods preserved in lye, which is a caustic chemical used to clean my floors. Come to think of it, I might need to really apply the same standard to my once a week soda habit.
The reality is, at all points in human history, we have been eating foods that are not ideal for our health and wellbeing. Part of this journey for me comes from figuring out where that line is.