It depends what institution you are at and how much research funding you have. If you are a professor at a small teaching university (even in the US), a few publications per year in open access journals is the same cost as supporting a student for a year. Depending on the field and how established you, obtaining research funding can take a lot of time that might be better spent doing actual science. (This is true at all levels, but in high-profile fields, the grant sizes are large enough that publishing costs are almost negligible.)

Although I am friends with David (the author of this blog post), I do not agree that open access is fundamentally the wrong model, but his argument is not without merit.