While you might not like to hear it, conventional psychology says that any behavior you take is likely to be repeated and that any behavior observed is very close to identical as doing it yourself (You remember it differently of course but brain activity is very similar).
So basically this should have been assumed but of course it's always nice to have the research to back it up... especially in the case where in games the characters are clearly not real people, of course once again psychology says we think in symbols so...
Anyway, all this means is that the research supports what has been the basic assumption and while the small change might not seem like a big difference when you think of it in a feeling, if you think of it as a group of action then it can be. If there's roughly a 2% increase in aggression, that might seem like the person is just "slightly more aggressive"... or probably should be right that that's an extra moment out of every 50 where the person becomes aggressive. Instead of the person laughing off that joke, they might get pissed off at you.
Just food for thought. Personally I don't think it needs the kind of stigma associated with such a warning. The government has no right to tell people who they may react to a given stimulus. Warning somebody that something might kill them is a whole lot different than warning people that they may kill someone.