Yes, abductive inference or some form of analogical thinking are how powerful hypotheses are really generated. Neither of the posts linked to in number 4 above even mention induction, so I'm not sure why the author thought they were evidence for the thesis.