Experimental evidence for very long term processes

By appealing to the Texas sharpshooter fallacy, you are assuming that there always exist a multitude of evolutionary pathways to deal with a given problem. But what are the evidence that this is true? In fact, it seems that the available evidence we have suggest the opposite, ie that very few solutions exist.