Let me add that we are talking about two different definitions of "fringe" here. This is, in part, a dispute over semantics.
Cude is quite right about what he calls "fringe" and I agree that is a valid use of the word. He is right that cold fusion fits that definition. However, I think that in the context of a scientific discovery, when we invoke concepts such as "fringe" or "marginal" or "proven" we should use the more rigorous definitions. We should stick to mathematical rather than popular culture definitions. When we talk about movies or politics, "fringe" is defined by whatever the majority thinks. Wikipedia or the New York Times are the arbiters. When we talk about calorimetry or tritium, opinions don't count. The majority view itself may be "fringe," even though that seems contradictory. The existing corpus of knowledge described in the textbooks sets the standard. Quantitative measurements such as signal to noise decide the issue. Not a headcount. Not who pulls political strings and gets to write Op Ed columns in Washington Post (Robert Park), or which anonymous nitwit named after a comic-book character prevails in the edit wars at Wikipedia. Decades from now, all knowledge of cold fusion may be lost. After I and others who know the facts die, the mythology alone may survive. The only references in textbooks or the mass media may claim that cold fusion was pathological science that was never replicated, etc. The Wikipedia/Sci. Am. version of history may prevail, because winners write history books. However, the Wikipedia version is incorrect. We can determine this by objective, absolute, universal standards. Cold fusion exists. It always has. It always will. Science does settle some issues beyond question. It is rather quaint to assert absolute faith in the scientific method, but I assert it! I may be mired in the 19th century, but I say there will never be any way to disprove the heat beyond the limits of chemistry, tritium and helium. Replicated experiments are the only standard of truth. Once you achieve a certain level of replication, there is zero chance the results are a mistake. Theory can always be overthrown. Experiments may be re-interpreted. But in this case, the results are too simple and clear-cut to be re-interpreted much. If the term "nuclear" means anything, and the distinction between chemistry (changes in electron bonds) versus nuclear (changes to the nucleus) mean anything, then cold fusion is a nuclear reaction, by definition. - Jed