This doesn't actually make the finding illegitimate however, though it does indicate we should use some scrutiny.
First, as I'm sure many of my older readers have already wondered, this study only focused on people who have been married at most since 2005. Given some lead time for publication and all, that means that they were studying the incidence of divorce in marriages in the first 7 years or so. Now this isn't totally crazy...about half of all divorces occur in the first ten years of marriage (This is what I learned in school, but now I can't find a good source for this, but this article seems to back me up), so this study does likely tell us something. It's interesting though that the abstract uses the word "slightly" to describe the lower divorce rate/marital satisfaction. It turns out that's pretty true, as the divorce rate for those meeting online is about 6%, and for those not meeting online it's 7.7%. This difference was smaller when they controlled for other factors, but was still statistically significant (they don't list it).
Now I don't think this is totally crazy. It's a small difference, but I would imagine that much of that could be attributed to people who went online looking for love/relationships vs people in the offline world who just fell in to relationships with people they encountered. Actively desiring marriage would, I presume have a protective effect on said marriage once it occurs.
Overall though, it is interesting to ponder where this might go. Are the divorce rates going to be higher once we get more than 7 years out? Are there other changes coming due to online meetings that we haven't noticed yet? Additionally, there's evidence that the divorce rate is not continuing to climb because many who would have gotten divorced are simply not getting married. As those folks continue to opt out, how will things change? I will be anxiously awaiting the eHarmony followup.