It's an interesting question why arguments from evil have become such a major family of arguments. I think it is a combination of two factors: (1) The relatively small logical space for positive arguments for the conclusion that God does not exist; and (2) the rise of the design argument.
To which he responds,
Why not also add the obvious facts that (3) evil is such a widespread phenomena and (4) it has great rhetorical and experiential weight?
And I think this is a great question, worth answering directly. To (4) there's a simple enough answer: precisely the puzzle is why it has such great rhetorical and experiential weight as an argument against God's existence. Certainly evils of various kinds are easy to find, and so we have a lot of experience with them. But lots of people throughout history have experienced evil without taking it as a reason for atheism. Lots of people still do, in fact. So arguments from evil to atheistic conclusions are not getting their rhetorical force from the mere experiential weight of evil; the puzzle is precisely to explain why evil has such great rhetorical and experiential weight in this particular context, given that it hasn't always had it. In explaining the popularity of the design argument, for instance, it would not really answer any questions to say that one of the factors is the great rhetorical and experiential weight of design; design does have great rhetorical and experiential weight, but this is in a sense what is supposed to be explained. And to explain the great rhetorical and experiential weight of evil for this particular sort of argument, especially compared to other arguments that in other places and times have sometimes been regarded as more forceful, we have to look at the conditions underlying the popularity of the relatively weak design argument, which is what sparked the popularity of the argument from evil as a response. I would suggest that these are, to put it in a rough and crude nutshell, a weak epistemology (thus reducing the appeal of 'ontological' arguments) and a weak theory of causation (thus reducing the appeal of 'cosmological' arguments); these same conditions also help to give the relatively weak argument from evil an edge over more robust alternatives. And, of course, each feeds off the other.
With regard to (3) I think we have to keep in mind that 'evil' is not a univocal word, and very few if any arguments from evil actually build on every kind of evil. 'Evil' in this context just means any sort of empirically discernible defect. So it is true enough that things we call evil are easily found; but it does not follow that the sort of evil that is considered in any particular argument from evil is a widespread phenomenon -- or, indeed, even noncontroversially evil or defective. Outside of clear moral evil, what you even count as a defect or lack of design may depend on whether you are already accept the design argument or the argument from evil in other cases -- someone who is elsewhere already a proponent of the latter may regard something as an obvious defect, whereas someone who already accepts the design argument elsewhere may classify it as a case where design is not particularly obvious; and vice versa. So how much of a role (3) can play in explaining the popularity of an argument from evil will vary considerably from case to case.
(He makes a few more interesting comments that are useful for clarification. I've come to think that the distinction between logical and evidential forms of the argument from evil, taken as classes of argument, is relatively unimportant. There is indeed a formal difference, but even formally it is a relatively trivial one: the difference is simply between arguments that are put forward to show that a conclusion is right and arguments that are put forward to show that a conclusion is more reasonable than the alternative, and that's really it. This is obscured by the fact that when 'the logical argument from evil' is discussed, it's often put in a very crude form, whereas the more common evidential forms are given more bells and whistles, but this is merely a sort of historical accident: if you wished, you could take arguments of one kind and with only slight modification make them arguments of the other kind. Every logical argument from evil has an evidential twin and vice versa; although, of course, an argument's plausibility doesn't guarantee the plausibility of its twin.
Also, I don't think the Fifth Way is a design argument. In Thomas Aquinas's thought a final cause is that which picks out the effect that an efficient cause will have; it need not be anything recognizable as design. The Fifth Way is a causal argument rather than a design argument -- indeed, I think it's intended to be an argument about the very possibility of causation itself, the very possibility of anything being disposed to have this effect rather than any other. St. Thomas does make use of design-like considerations here and there, but as one would expect they chiefly occur in discussions of divine providence.
And, should it not have been clear enough, I myself think that at least some 'cosmological' arguments still have force.)
