This paper in Frontiers of Ecology and Environment by Andrew Zumkehr and Elliott Campbell conducts a type of simulation to suggest that most Americans could eat locally (see the accompanying press release here).
That seems like the wrong question. It shouldn't be whether we CAN eat locally but whether, WHY would you want to eat everything grown locally?
The paper basically assumes local is "good" and asks how do we get more of it. A few, uncritical references are made to the fact that local food systems may "result in large GHG emission reductions" or that it "shortens the distances required for economic and energy-efficient recycling of waste streams between farms and cities" or it " may increase community involvement in food production issues, potentially leading to improved environmental constraints on landuse practices." Yet, there is good reason to believe that local food systems would generate more GHG emissions, result in less food choice and dietary diversity, increase price and availability risk for consumers, and drive up food costs. There is a lot written on each of these issues, none of which is referenced here.
In any event, the authors do some calculations to suggest it may be technological feasible to provide enough calories to feed everyone in this country with local production. The authors used yield data from each county in the US to infer the productivity of growing crops in each location. However, it is likely a mistake to assume that yield would remain constant as production expands to more marginal lands. In fact, it is almost certainly the case that observed yields near urban locations are an upper bound for the productivity in the area because only those lands that are currently productive enough to out compete other uses are those currently in use for crop production. That is, you're only observing yield from the most productive lands and you're not observing yields from the least productive lands.
It might not be surprising to hear that the authors don't calculate the cost of all this. The words "cost" and "price" appear exactly three times in the paper, the latter of which in reference to the fact that they don't study price effects. The paper concludes that:
“current foodshed potential of most US cities is not limited by current agronomic capacity or demographics to any great extent, and that the critical barriers to this transition will be social and economic.”
Saying the main barrier is "economic" is akin to saying the main barrier is reality. The reality of the resource constraints that nature deals us and our willingness to pay to overcome some of those constrains.
Nonetheless, that doesn't keep them from proposing some grand plans . From the press release:
“Campbell’s maps suggest careful planning and policies are needed to protect farmland from suburbanization and encourage local farming for the future.”
I don't see anything in this paper that suggests we need "careful planning" or to "encourage local farming." If people want local foods and are fully willing to pay for them, farmers will provide it.