Matt summarized the end, but the problem with the article is that it compares the wrong things.
-DOF is related to the distance to the target. Long distance, long DOF. Getting a full body shot with a 50mm means you are long ways off. Not the ideal use of the lens. In that case the background is also disproportionately close minimizing bokeh.
-The compared 1.2 to 1.4 to 1.8. Should compare 1.2 to 1.2, 1.4 to 1.4 and 1.8 to 1.8. Then they jump to F2 to compare. Almost all lenses are a bit soft shot wide open. Of course the further from center the softer they are (by definition wide open means you are using the whole spherical portion of the optic). Closing that optic down and getting closer to the center of it will ALWAYS sharpen it. That being said more curvature being exposed (1.2 most) will also always give you the best bokeh.
-They forgot to worry about chromatic aberration (typically worse wide open) as this is also from the non-center of the lens. The 1.4 is much better than the 1.8 here
-They also ignored lens flare. Often with in situ portraits there will be bright lights around you. The 1.4 deals with flare significantly better than the 1.8.
-They also did this on a full frame. The 1.4 is completely different on a crop as much of the outer portion isn't used.
The whole article is basically crap. Problem with a ton of them on the net. Almost like they expect their readers to not think. Drives me nuts.