I wonder about how the end-to-end efficiency of this compares to going back to older processes from before synthetic nitrogen fixing.
Reading between the lines, they are fixing nitrogen by converting it to ammonium nitrate using the Haber and Ostwald processes, with energy the run the process provided by solar panels, and then scattering the ammonium nitrate on the soil where they grow their crops.
The classic way to fix nitrogen is to grow nitrogen fixing crops sometimes (members of the Fabaceae / pea family), such as peas, clover etc..., which fix nitrogen, and rotate the crops growing in each plot of soil over different growing seasons between crops that deplete nitrogen and crops that fix nitrogen. It's entirely possible that this is more efficient in terms of land use (area * years) than using some of your land for solar panels and then growing the crop you want continuously in the other part of the land. It also means lower embodied costs in terms of resources to manufacture the solar panels.
So it would be good to see some actual numbers around this.