That's a pretty cool pictorial comparison of things for sure. It brings to mind a bunch of similar pictures from many many years ago.

The Precision Aerobatics guys from the dim past spent some years discussing aft airflow and how to control it. They didn't stop with the trailing edge. Well, actually, they started slightly ahead of it.

Some of them noticed the aileron/flap/elevator/rudder gap. As with everything, the line drawn created two different camps. A bunch of guys decided the TE wasn't really as important as the other side thought because the gap turbulated the air before it hit the active surface, much less before it hit the TE. Then LOW AND BEHOLD, a bunch of guys decided that every wing profile developed up 'till then showed the air turbulated before it was even close to the gap.

Then a bunch of guys started screwing around trying different ideas and did their best to explain it all with aeronautical terminology. Some guys "discovered" that if they made the leading edge of the surface significantly wider than the trailing edge of the wing, their models flew awesome. Of course anothe bunch "discovered" that making the LE of the surface significantly narrower than the TE of the wing, THEIR modesl flew even awesomer. Good thing was both bunches had the same theory. The suddenly wider LE, and the suddenly narrower LE both acted to reattach the boundary layer or somesuch.

All the while, a smaller, less vocal bunch found that making the surface's TE both square AND narrow made their models fly more accurately, faster, more efficiently and many other awesome things that have since slipped my mind.

I believe there has been some relatively scientific documentation that somewhat proves the square and narrow TE treatment. One thing for sure about it, it's easier for a modeler to create accurately, and far easier to tell when it's gotten hangar rash.

One problem us modelers have that affects the whole situation is accuracy. I maiden a bunch of models all year long. One thing that is almost guaranteed is that almost every one will have at least one surface hinged inaccurately. I'd put $20 on a bet the models in question are going to be affected more by less than perfect alignment of active surfaces to TEs than by anything else. But discussing theory is fun for sure. The beauty of modeling is we so very seldom have any really accurate ways to measure our results.