In stating that the two events were independent, I was assuming that the gender of the second child was yet to be determined. Things change though when both genders are determined in advance. The 1/3 chance makes sense now, as well as the connection to Monty Hall.

Thanks for arguing the articles case, it helped me learn something. Namely, to read the problem more carefully to determine the assumptions correctly.