How is the stress at a probe location obtained in Simulation 10.0? You can insert a probe in Simulation 10.0 in various ways; either by clicking on an arbitrary location with your mouse, clicking on a vertex or defining it based on the origin of a coordinate system. When you solve your problem afterwards, how will the probes follow the deformation and how are the results evaluated at the probes?
If you define your probe on an arbitrary position, for instance between two nodes in a two dimensional model, will the probe follow the deformation of the 2 nodes on each side and will the results for the probe be an interpolation between the node results on each side of the probe? Or is the probe position fixed in space so that the result will be more and more inaccurate as the deformation is increased?
I have done some tests and my conclusion is that if you define your probe to a vertex the result will be evaluated on the vertex since WB always puts a node to a vertex. So when the node is displaced, the probe will give a correct result for the displaced node location.
As you observed, the stress at a vertex will be the stress at a node and will not be fixed in space. The stress at a coordinate system will use an interpolated value based on the original position of the model, so it will move with the nodes used in that interpolation calculation. There is an older, less accurate probe which provides a fast estimate of a contour value at an (x,y,z) location when you click on it and drag the cursor to the desired spot on the contour plot. WB then draws a temporary annotation of the estimated contour value at the cursor`s location based on the color of the pixel at that location. This is a very fast calculation but is not as accurate as the probe associated with a coordinate system. 

