Selection bias alert (porn is bad for you edition)!

One thing that consistently irritates me about most mainstream journalism is that journalists are seemingly untrained in basic logic or statistics. They will unquestioningly relate any opinion expressed by an "expert," no matter how unsubstantiated or flimsy. The fictitious headline, "New Evidence World is Round: Experts Divided" expresses this tendency well.
And so it is in an AP article entitled Battle Brews As Porn Moves Into Mainstream. The battle, apparently, is between those who think that pornography leads to addiction and inability to maintain a relationship and those who do not. As the author David Crary writes, "The bottom line, perhaps, is that each side in the debate can make points that seem unassailable." This is typical look-at-me-I'm-so-objective journalism but, I believe, does a disservice to readers.
The evidence cited that porn destroys relationships is that for up to 10% of porn consumers, "relationships suffer." Even taking this statistic at face value, it's rather like saying that drinking bathtub gin makes you poorer since consumers of bathtub gin have lower incomes than those who consume Scotch whiskey. The notion that people might become (self-select in stats jargon) porn consumers because they are in failed relationships does not occur to Mr. Crary nor to the "experts" he quotes. I had to laugh out loud when I read the quote, "many husbands spend so much time online that they cease to have sex with their wives." Hmm. Do you suppose there is another explanation for this no doubt interesting phenomenon?