> Yet it is very successful in all its predictions. So obviously the Popperian "falsifiability" is too simple a criteria for science.
Those two sentences contradict each other. If something unobservable implies some observable, i.e. predictable, consequences, then falsifying these predictions would also falsify the unobservable antecedents.
The hidden variable/better (unknown) alternative theory problem is what he is referring to.
Newtonian Mechanics are falsifiable, but no one could figure out an example that actually did so, until roughly when the orbit of Mercury was fully observed.
Those two sentences contradict each other. If something unobservable implies some observable, i.e. predictable, consequences, then falsifying these predictions would also falsify the unobservable antecedents.