Hang on the survey was done on people using the service for Married individuals to discretely meet other married individuals...
For a start that means the survey was only run on those cheating.
Which means very little in terms of the outcome of the survey.
What it really says is that of those people using the site. They are twice as likely to drive a BMW.
But that doesnt look at the overall amount of BMWs in the population as a whole.
If for example the population had twice as many BMWs as any other brand on the road then your seeing no difference between those who are cheating and the overall population.
Its pretty much a nonstarter of a survey as its not a fair representation of the population, being only for Married people, with an internet connection, who are brazen enough to use a website to cheat.