Gibbs is incorrigible. And ignorant. A couple of months ago he wrote a column in which he tried to explain some basic theory of science concepts such as: the difference between theory, hypothesis and model, and the meaning of experimental evidence and replication. I don't recall the date of that column. I do not think we discussed it here. Anyway, he was hopelessly confused. He does not even have an undergraduate level grasp of these things. He used words like "theory" in the colloquial sense, referring to what in science is called a hypothesis. Or a wild guess.
I have ignored Gibbs since he wrote that. I did not even bother to respond in his remarks space. He is another "science writer" like Lemonick of Time magazine who would fail an undergraduate introduction to science course. He is unqualified. It is as if he is writing for PC Magazine and he does not know the difference between firmware, operating system software, and applications software. (In the early 1980s I read an expensive executive report on the "Future of Computing" by a vaunted industry expert who did not understand this difference.) There is nothing wrong with using the word "theory" to mean "hypothesis," in the proper context. When you talk about a police investigation or why a programmer thinks a procedure is giving the wrong answer, it is fine to refer to a "theory" -- meaning we don't know and we are going to test the idea, or look for supporting evidence. The problem arises when you mix up different meanings of the word from different contexts, for example when people who know nothing about biology assert that the theory of evolution is only a hypothesis, and that it is seriously challenged by competing hypotheses. As it happens, theories about cold fusion are also hypotheses. Some should probably be called "wild guesses" or vaporware. However, suppose in the future one theory emerges, becomes widely accepted, and is shown to predict the behavior with great accuracy. Suppose it eliminates the competition. It will still be a theory. Theories never graduate to become laws. Actually, in my opinion, a law is somewhat less than a theory, being an unsupported single repeated observation without any deeper known implications. As far as I know, Newton's formula F = G*(m1*m2/r2) is still the law of gravity. It is not a theory because no one knows why it works. There is no deeper implication. (Okay, maybe there is but I haven't heard about it. I realize efforts are underway to detect gravity waves.) - Jed