Well he's better than most on this issue.
A much better study would be to have a control group and a playing group. Both groups can't drive for a week. During this week, the playing group plays Sleeping Dogs and just drives around the map on the left side of the road. This is in third person view with a gamepad.
After the week is up they through them in real cars and just have them drive around the block. The test subjects then fill out a simple survey including how they'd rate their desire to drive on the left.
If the test group has a higher tendency to desire driving on the wrong side than the control group, then you have an actual study showing the effects of in game actions on life outside the simulation.







