Nikki, the brace works in conjunction with the triple tree and axle.
Please see the image below. This image ignores the triangulation which exists by nature in a tube and clamp situation. Since both ways have that equally, it is a null factor in this discussion:
Figure a and b ignore triple tree and axle rigidity, since they exist equally on both setups.
Figure A is without a brace. The only structural rigidity are the tubes themselves. In figure b, we see the triangulation via a form of gusseting that occurs with the brace. I'm sure I don't have to explain the benefits of triangulation. I haven't added all the lines for triangulation, because it decreases greatly due to distance between, say, top of left stanchion to the two mounting points on the right fork lower, and the design of the stick brace isn't really the best at dealing with that direction of force in the first place
The other 'group' benefit can be shown in c and d.
Boxes aren't the most structurally sound shape. of course, the longer one side of a box there is, the more leverage a force has on it. By adding the brace, we reduce the size of the boxes, and reduce that leverage.
Again, this is simplified. I don't have my cad program installed here at work, and drawing it all in, basically, mspaint, would take too long.
Hope this helps explain why your tests are not indicative of actual benefit, and hope it helps people understand more why a brace does help.