This is a tricky section. I think that Hume wanted to criticize all arguments from design. That he approached his quarry so indirectly is either evidence that he was afraid or that he thought this was the best way of making his argumentative medicine go down.
What’s indirect about what he wrote? Well, most of the section operates under the assumption that a basic form of the argument from design is true. The wonderful and complex features of the world show that there is an intelligent creator who made them. We can legitimately infer that an intelligent designer is the cause from observing the effect (the wonderful and complex features of the world).
The arguments are officially directed at two discrete uses of arguments from design. These move from features of the observed world to conclusions about God’s providence and justice after death. (The latter is especially relevant to the matter of whether atheists pose a threat to public order and morality since the prospects of divine punishments and rewards are powerful incentives for good behavior.)
Plus, the arguments are put in someone else’s mouth, twice over. Hume does come clean in the end, however. The last paragraph, which gives the real point of the section, is unambiguously his.
We saw how Hume’s discussion of miracles throws light on his skepticism about inductive (a.k.a. causal) inferences. Hume took induction seriously and held that there is a big difference between our reasons for believing in the uniform laws of nature and our reasons for believing in miraculous exceptions to those laws.
It’s a nice question whether he can reconcile that position with his skepticism. According to his skepticism, we have no more reason for thinking that a regularity that we have uniformly observed in the past will repeat itself in the future than we have for thinking it will not. My only point here is that he tried to reconcile the two positions.
By the same token, our process of inductive inferences is taken quite seriously in this section. Hume rejects the argument from design on the grounds that the causal inference it makes is different than our normal ones. That doesn’t make any sense unless our normal causal inferences make some kind of sense.
We closed by mentioning one of Hume’s recurrent themes. Philosophical speculation is harmless whereas religious speculations leads to undesirable enthusiasm, where ‘enthusiasm’ means something like fanatacism.
I said that I didn’t know whether religion has malign consequences but that Hume could not have gotten away with his blanket statements about non-religious philosophy today.
As it happens, Scientific American has weighed in on the question of whether religion a necessary component of social health. The answer: the data are conflicting.