When somebody does some objective, rigorously controlled tests and publishes the results in a reputable, peer reviewed journal they'll have my attention. Until then it's all hearsay, subjective and anecdotal. Great for a bit of entertainment but that's about all.
While I agree that I would like to see credible data, I actually don't doubt the theory.
Several years ago, when we switched the electronics industry from lead/tin solder to lead-free solder to apease the environmentalist whackos (don't get me started) one of the huge concerns was -- and remains -- the long-term reliability of lead-free solder joints. The pads on a printed ciruit board to which the components are soldered are made of copper. But copper oxidizes quickly when exposed to air. To protect it, we have, in the past coated those pads with the same lead/tin alloy which would be the solder; that worked perfectly for the industry for upwards of seventy years. We could coat the exposed copper with just tin, but tin has a problem: over time, it grows these little "whiskers" which, given today's microscopic component spacing, can cause short circuits. Also, while it's not as bad as bare copper, tin oxidizes. (By the way, the growth of whiskers in tin is well-known though only recently explained, and it should be enough to convince you that metals do change over time.) One solution is to plate nickle onto the copper. But nickel is very difficult to solder to, so you have to then gold on top of the nickle. So, now the pad actually consists of copper with nickle on top of it with gold on top of that. And we're now going to solder to it with a solder which is tin, silver, and a pinch of copper. Wow! We now have copper, nickle, gold, tin, and silver all in one place. Think of all the possible reactions which might occure!
But, it's not so bad because, for example, the nickle won't ever interact with the tin in the solder because there's a layer of gold between them. Or will it? It actually will because atoms of metals can actually move through other metals. So, slowly, over time, atoms of the nickle will move through the gold and interact with the tin and the siver. Slowly, over time, complex intermetalic compounds are created. Intermetalics have may properties, but most of them tend to be brittle. So, over time, the solder joint will get brittle and can then crack and break. Accellerated life testing has already shown this to happen.
My point here is that metal is not static. Inside of a seemingly unchanging piece of metal, there's a lot going on.