This is nonsense, because this is the standard fallacy of any survey. You take a sample and extrapolate out to the general population. If your sample is not representative, then the statement is false.

In fact, I would go so far as to say that we should mandate a law whereby nobody can claim "% of X agree to Y" unless it is specifically stated "% of X surveyed out of a sample size of N agree to Y. Here's a link to the methodology and questions."

Organ donation is good and important, but default opt-in is a bad decision, and the reasoning listed for it is worse.