You points are all valid and definitely worth considering. My issue with circumstantial evidence is that it can sometimes lead you to the wrong conclusions. For instance, people were convinced the 120 patient trial was done and we would hear the results any day. Then, on Wednesday we were told the 120th patient has just been signed up.
Perhaps I'm being a wee bit too cautious, but as we used to say in my I.T. days "in God we trust, all others bring data".