While it doesn't invalidate the main points of the article, we should be wary of comparing countries that produce trustworthy data, like Japan, Germany, Italy, Spain, Korea, or France, with countries that don't. The reasons national stats from a country can be unreliable include the lack of technical means to collect the data or the absence of a free press. 

Even in the US, inaccuracies creep up when the numbers have political or economic consequences. It is well known that Americans are undercounted because politicians won't allow the use of advanced statistics to detect and correct known undercounts. I don't think we know how many people actually live in China. In another country I visited long ago, an attempted census in a province had yielded a count "between 200,000 and 1,200,000." Some parents had declared more children than they had in hope of getting government subsidies, while others had declared fewer, to avoid taxation. The census takers had no idea how many of each there were.