On symmetry‐modulated distributions: revisiting an old result and a step further