The thing is, pathologically evil people and insane people have, at times, made important discoveries. They have done first-rate research.
Fair enough. But, if we are to do "likelihood from character" Rossi is neither evil nor insane, merely much much more interested in dramatic effect than his stuff actually working. To the extent that he has always said tests that might establish that are "not necessary".
Or, if we do likelihood from ability, training, etc, Rossi scores zero.
So, while you could I suppose claim that anyone could make a serendipitous world-shattering discovery, Rossi looks a supremely unlikely candidate.
His only qualification would appear to be a good opinion of himself and the ability to convince others of that.