I don't think the argument is sound. If I walk into a big room with the general assembly of humankind and shake hands with a random person, that person will probably be from China, Russia, US, India or some such because those few countries have more than 50% of Earth's population. However if I walk into the UN where each state has one seat no matter how small and I shake hands with a random person, that delegate will likely be from a small country because there are many more small countries than big ones.

Now if I were to resettle humankind to 250 or so exoplanets and gave each country its own planet, and each planetary civilization eventually brings forth a smart guy who reasons along the lines presented in the Big Alien piece, then I end up with 250 papers or so arguing that surely most exoplanets will have a much smaller population than their own. Almost all the papers will have guessed wrong.

Note that while one can always ruin likelihoods by tampering with the data, in this case the exoplanetary populations were modeled on a realistic sample that does follow a power (Zipf, Pareto) law.