With all due respect, that's not much of a response in a serious discussion. Can you elaborate on what the physical phenomena are that are playing a role in your hypothesis?
If you are onto a worthwhile model for stress analysis of welded joints, I for one would like to learn more about it. A complete response should include definition of the parent material, filler material, temperature-time history of the filler & adjacent material, welding process (TIG, MIG, stick... ?), description of the as-welded bead characteristics, assumptions about the grinding/smoothing process, a physical explanation for why the smoothing process weakens the joint, and an explanation for how bending at 90 degrees is relevant to the design conditions for the joint in question. Without this information, I maintain that your original assertion is merely a generalization without qualification.