Things aren't that simple. The early transistors weren't reproducible...not predictably. And nobody knew why. It eventually turned out that they could be poisoned by trace amount of materials below the amounts chemically detectable at the time. IIRC it took over a decade of very careful work to figure that our, or it may just have been to figure out how to prevent the poisoning. And that had significant money behind it. (I think it was pre-breakup AT&T.)
Now I haven't seen anything convincing that indicates that cold fusion will work, but I also haven't heard of any significant investigation. Merely various spot checks by people who say either they can't get it to work or "I'll sell you this black box.". I'm dubious about its actually working, but not convinced, and don't see any reason that anyone else is convinced...either way.
To me this seems like "this is a low probability proposal which has some claimed marginal evidence and no reasonable theoretical justification and no convincing evidence". Remember just how difficult it is to actually prove that something is false, where you don't know care what mechanism that might be causing it. Were I investing, I don't think I'd invest in it, because even though the potential payoff is astronomical, the probability is extremely small, and the difficulty in reaching a definite negative proof is extreme. And other people have already failed to reach a positive proof. And only a positive proof has a reasonable payoff. (Buying Lockheed stock seems like a better use of the money.)