Choice B is the best answer. Chart 1 shows the percentage of the US population in 2010 that lived in non-metro, small metro, and large metro areas. While the author cites census numbers, he notes that "when it comes to measuring demographic inversion, raw census numbers are an ineffective blunt instrument" (lines 11-13,"But... instrument"). Census data refer to the number of people living in a specific area and the demographic information that's been collected on them. The author would most likely consider the information in chart 1 to be possibly accurate but an "ineffective blunt instrument" that's not truly informative.
Choices A and C are incorrect because the author would not consider census data to be excellent or compelling. Choice D is incorrect because while the author does not believe the census completely explains demographic inversion, he would be unlikely to disagree with the census data.