I would have thought it obvious: it looks like a lateral thinking puzzle to some people, and it's painfully common to have the null case be a part of the solution (in a sort of perverted aha-gotcha! attempt).
After all, part of the linked article's point was how it struck him as potentially broken; if the problem seems broken, then many will heuristically start attacking the problem itself rather than the math. And once the problem's no longer trusted to be stated as exact and correct, well, a simple pedantic jump in meaning is a pretty small step. (I'll concede it's possible you've never come across a broken 'clever' math word problem before, but having found enough to be wary myself - both in school and college - I'm rarely surprised at people leaping for the lateral thinking solution.)