Depending on what sort of city driving you are doing, the fact that the porsche has higher T levels in the city than the camry does in the country says something. The article argues it's the status effect. (I wonder though - if city driving means stopping (and then accelerating away from) traffic lights. That would be fun in a porsche - boring in a camry.) I'd like to see the test done with an accelerometer in the car. (But us armchair scientists can always find something wrong with an experiment can't we?)
On a related note - I really miss my old 1990 Toyota Camry - I used to be able to drive it nice and fast on some of the twisty dirt roads around here. Didn't really care what happened to it in terms of bumps, scratches, stone chips. I wouldn't do that in a 150K porsche.
Maybe the drivers DO feel more manly in a porsche because the drive is far different, which it is.
To draw a more accurate conclusion as to why, it may help to know how long after driving the increased testosterone levels lasted for.
If you really wanted to figure out why this occurs, a better study could involve a porsche fitted with a Toyota engine, and vice versa, to provide the same drive quality but equivalent external appearance.
Might also help to have a Porsche with a Toyota engine in a simulated environment where all other drivers are slower (so the Porsche driver is still relatively faster than other cars). Maybe the effect is from "dominating" other drivers.
TL;DR: Any speculation on why this occurs is pretty much baseless based on the information provided in the article.