In the first post, I gave some indications of why I think denying God's foreknowledge of free human acts doesn't really explain that much evil. What I'd like to do now is lay out a number of elements of the traditional response to the problem of evil, the one that open theists find unsatisfying. This will all be at a fairly basic level, but I'd like to get all the general things on the table before going into depth on how denying foreknowledge is supposed to help.
One of the primary strategies for responding to the problem of evil is to treat some good as a higher-order good in the sense that it can't exist without allowing some evil to exist yet the good is worth the evil it allows in some sense. Many traditional presentations of the problem of evil have assumed utilitarianism, and thus they will talk about the consequences for happiness and unhappiness, saying that more unhappiness is created than the happiness that requires it, so it's not ultimately worth it. Some theists have responded that utilitarianism is false, and thus the theist has more resources to explain evil. Some kinds of evil may simply be wrong to prevent, with no relevant questions about how much evil is allowed by not doing that wrong thing. If it's wrong to do it, then God shouldn't be expected to do it. So I don't want to assume utilitarianism here, even though it's easier to frame the problem of evil if you do have such assumptions. The way to think of higher-order goods in a non-utilitarian framework would be to see some goods as being so important that it would be wrong not to pursue them. Alternatively, one might simply see preventing certain evils as morally wrong, because any method of preventing that kind of evil would involve doing something wrong. Most theodicies or defenses (I'm not going to deal with the distinction some philosophers make between the two) fall under some kind of higher-order good, I would say.
The most common higher-order good is human freedom. Plantinga's free-will defense importantly shows that it's possible for human freedom to be so valuable to God that the evil it allows is worth allowing. A non-utilitarian would describe this as its being wrong to prevent human freedom for the sake of preventing evil. Since open theists don't disagree with this, I'm going to grant it as part of the traditional picture that is common with the open theistic picture and thus refrain from dealing with it any more than just this summary.
Another higher-order good would be soul-making. Soul-making theodicies include personal growth through suffering. If suffering can lead to greater spiritual progress, it might be worth it. It's harder to frame this in a non-utilitarian way, since it doesn't seem wrong to prevent suffering that might cause someone to grow. If, however, growth is impossible without suffering then it seems as if you have to compare which is more important -- the growth or the prevention of suffering. When no other considerations enter in, even someone who isn't a utilitarian might use utilitarian reasoning. It's perhaps because I'm not a utilitarian that I never understood the soul-making theodicy except as a supplementary reason given that there has already been a fall and a need for growth away from a fallen human nature. Once you have that situation (which the free will theodicy provides), it falls right into place. I should mention that Christians have some more elements to this than other theists, since identifying with Christ in his sufferings is something the apostles insisted on. Through suffering for the sake of Christ, believers experience something of his own suffering and thus identify with him. I wouldn't place this on the top of any list, simply because if there were no evil then there'd be no suffering of Christ and no need to identify with his suffering, but once these other things are in place it seems this is perfectly fine for Christians to add as an explanation of some suffering Christians endure as a result of their being followers of Christ.
Natural laws create another set of important higher-order goods. All sorts of things are required for real relationships. Human freedom is just one. If one of God's primary purposes in creation is to have beings who will be in relationship with him and with each other, then certain things need to be true. They will need to be able to predict the consequences of their actions, else they'll try to do something loving to someone, and it will lead to pain. So we need natural laws. Now it's possible, as far as I can tell, that a complete set of natural laws will enable people to predict their actions without ever allowing anything dangerous, but it's also possible as far as I can tell that there is no such set of laws.
Natural consequences are also important for most theists. If God disapproves of evil actions, then wouldn't it make sense for God to create situations that will discourage those evil actions? This would explain why there should be negative consequences of evil actions, why those who resist God's good plan for wholesome and healthy relationships will tend to face negative consequences. If God's goal ultimately is for people to be restored to a relationship with God and with each other in ways that overcome the negative tendencies of the fall, then there needs to be a way for people to see how bad our plight really is, to see that we need God's help in getting out of our situation. Otherwise, we may well just do it again if God were to fix things immediately and remove our memory (which would be required if removing hate were to be a part of it). Also, genuine relationships require genuine interaction and not just a personal virtual reality like the Matrix for each person to prevent harm to other people when we do evil things.
The main response from atheists who think the problem of evil is a good reason not to believe in God is that these explanations of why God might allow evil do not explain the amount or severity of evil in the world. It's true that they show that God's omnipotence, omniscience, and perfect goodness don't entail the nonexistence of evil, and so theists who believe all those things and admit to the existence of evil aren't running into a contradiction, as J.L. Mackie had once claimed. What still is true is that these explanations don't show why things have to be as bad as they are in terms of severity and extent. Why the Holocaust? Why American slavery? Why does God allow children to die grievous deaths at the hands of child molesters? Why can dictators like Saddam Hussein get away with running people feet first through a shredder so they can feel the pain all the way up? God could have eliminated some of this without sacrificing any of those higher-order goods. God could have reduced even the kinds of evil, preventing any of the worst kinds for the sake of allowing only what was necessary to achieve those purposes.
There are two ways to respond to this. Peter van Inwagen simply says what seems to me to amount to "God doesn't care". He starts from a point about vagueness. There may not be a sharp line between the amount of evil that will serve these purposes and the amount that is too much. So God has to make a decision somewhere in that fuzzy area, and if it ends up being higher than you would make it, too bad. Van Inwagen denies the Principle of Sufficient Reason, so he doesn't think God has to have a reason to pick point A rather than point B. One might object that point A involves more evil than point B but only still serves God's purposes as much as point B already does, so God should go with B over A. The problem is that you start a sorites series with each being slightly less evil and being similar in serving God's purposes, but eventually you've got to reach a point where it no longer serves God's purposes. The point of invoking vagueness here is to say that there is no lowest point of evil that still serves God's purposes. So far so good.
Then we get to the part that sounds to me as if he doesn't think God cares. Van Inwagen's further claim is that God doesn't really have to consider things like whether one person is inconvenienced more or less in a given day or whether someone has a headache rather than not having one. God has a general plan but doesn't include in it things on that scale, unless the headache has further implications of greater importance. This is the direction open theism heads when it says God's lack of foreknowledge explains evil. There are certain things in God's plan, and God will ensure those things, even if it means violating someone's free will (which presumably an open theist who affirms Christianity would have to say could have taken place with Judas to ensure that Jesus' betrayal take place). Other things aren't as important, and God doesn't work out all the details carefully enough to have to explain why he allows those.
How can this be better than the traditional approach? It's supposed to absolve God from having to have an explanation for as much evil, but it does so at the cost of making God an unfeeling, uncaring miser. Traditional theism says God really does care but doesn't prevent suffering. Open theism says God doesn't even care. How does that count as a response while retaining perfect goodness? I can't see how it does.
Compare the traditional response to this problem. It says that God does care, thus retaining perfect goodness. Both claim to retain omniscience, though open theists have to be of the sort that denies any future truths (at least about the things God doesn't know) to say this. The only threat here, then is omnipotence. If God knows and cares, then why doesn't he prevent these evils that van Inwagen says God doesn't care about? The traditional theist tries to have the aforementioned theodicies covering as much as possible. That much isn't too controversial. Some theodicists add other, more controversial, claims, such as Malebranche's insistence that the set of natural laws be as simple as possible consistent with not allowing too much evil, since the aesthetic value of a more perfectly-running universe should count as a higher-order good. Still, William Rowe is right to claim that there are many instances of evil for which the theist has not given an explanation, and there are degrees of evil that could have been less while all the aforementioned criteria will be met. This may be so. I'm not sure why the traditional theist can't give the first part of van Inwagen's response about vagueness to lessen the blow of this objection, but it doesn't say enough.
There's one crucial claim that theists have traditionally said that none of this pays much attention to, and that's the primary lesson of the book of Job. Rowe apparently thinks he can envision whole systems of natural laws about things physicists are barely beginning to understand and much more complicated, because they have to cover things at levels far below what we now know. Then he apparently can sort through all those systems to see that there are some that will be consistent with God's purposes but allow less evil than we have. Does he have a right to say that he can do such things? Absolutely not. What's worse is that we can't really be sure we know which things are most important to a perfectly good being who is omniscient (and therefore which things are most important period). There may be intrinsic goods that we can't know anything about due to the limits of being human. Rowe thinks we should have seen such things by now, but why would he think that?
We also have to consider things like how an afterlife or even centuries or millenia possibly of future even in this life would play a role in comparing the evils now with whatever is true over history. This is an easier calculation to make if you're a utilitarian, but it's a point that can play a role even if utilitarianism is false, especially if the two points of comparison are what I see now and what I will see to be true in the end. There are other things along these lines that I've seen pointed out in papers by William Alston, Daniel Howard-Snyder, and others, but one final element will suffice to make the point here. If the Christian story is correct, and a fall took place, a fall that corrupts our ability to estimate what's truly valuable and good, then why should we expect to be able to make uncorrupted judgments on such matters at all? Christianity, then, has a perfectly good explanation of why nonbelievers would think certain evils problematic for God to allow when the reality is seriously otherwise.
The conclusion of all this, on the traditional picture, is that God may well have perfectly good reasons for allowing as much evil of the kinds we have. Anyone asserting the contrary faces a strong argument. Rowe seems to think the theist has the burden of proof to show why there are such reasons, but it seems right to ask for such reasons only if we have reason to think that we would see such reasons if they were there. These arguments seems to go the other direction. It's not just that we don't have reason to think we'd see such reasons. We have reason to think we wouldn't see them. So Alston, Howard-Snyder, and others who take this Job line have a strong case against the argument from evil. Given that, I have trouble seeing the need to go further and explain more evils that result from lack of foreknowledge. It's not just that open theism doesn't do very well at explaining them, as the previous post argued. It doesn't even have the need to. The traditional view of God seems as good as you're going to need once you've made this move. So it's not just not sufficient. It's unnecessary.