I must disagree with your conclusion here. Probabliities are a very useful way of assessing the feasibility of a scientific model when experimental falsification is not possible. In fact, this is often done with biological, geological and cosmological evolution (as you mentioned.) The reason is that neither evolution (of any variety), ID or Creationism constitute scientific theories because none are subject to falsification. It is simply not possible to reconstruct all the conditions of our earth (or universe's) past in a controlled laboratory environment and subject the hypothesis to falsification. Consequently, none of these can truly be considered scientific theories, but rather models of origins.