Actually, just to see what's happening I've made a quite simple model: a box with an applied load in one side and a encastre in the other one.
I cut the box just in the middle, I plot the FreeBody Loads by components and, as expected, I see the resultant load in one of the axes that corresponds with the applied load and almost 0 in the others.
If I maintain the same point but I change the cutting angle, the values of the others two axes starts to change but the resultant value remain the same.