You get it wrong. It's their responsibility to show 0.8 is enough. I dunno if it works as a general counter argument in psychology, but in physics (and non-diseased disciplines in general) you do have to account for errors in your measurement apparatus as a possible source for your correlations, and if you don't, no-one is obligated to believe you.

edit: imagine you had been measuring speed via proxy that has at best 0.8 correlation with speed.

edit: okay, the article seem to find stronger correlation between belief in god and the failures at their test, than between belief in god, and IQ. At same time, there's many ways to describe what their test measure, e.g. intellectual laziness would do as well (the fallacious answer is a mathematical operation on last 2 numbers you heard - won't get simpler than this), or the arrogance (never self doubting enough to check if the answer makes sense).