Sorry this is kind of long, but I figured someone should try to help out the Theists.
I don't really agree with Prager, so much as I disagree with Harris.
His side of those emails seems to be a pretty clear case of him piggybacking one argument on another.
The first argument is that God doesn't exist. That argument is successful, so long as we take him to be saying that there is no empirical/scientific evidence to suggest the existence of God. As a scientist, that's all he can really say. Sorry for the Theists, but they're just gonna have to concede that to him.
His next step though is much more suspect. He goes from making a statement about what science tells us about the nature of the empirical world, to what we should or shouldn't believe in. At first, it may be hard to see why that is sort of sneaky. One the one hand, science sort of does that all the time. It tells me that I should believe that when I jump up that gravity will pull me back down. Harris' argument is different. His argument involves a moral "should" and is straightforwardly utilitarian in nature.
He seems to be relying on the argument that the world would be much better off without religion. That without religion, the world would be less violent, oppressive, and miserable all the time. Since he is scientist, I think it's fair that Harris should have to produce some kind of evidence to support that kind of claim, but I think it's unlikely he could find any. The best he could say is that there is a high correlation between groups of people who are violent and groups of people who are religious. There is a big difference between correlation and causation though. 'People with hearts' and 'people with kidneys' has an extremely high correlation, but people don't have hearts because they have kidneys (at least not in any sort of straightforward sense of 'because'). If he wanted to actually show that religiosity (is that word?) causes violence and conflict, he'd need to observe the whole history of a control group who were kept completely free of anything resembling theism.
Does anyone really think that such a group of people would really show a significantly lower incidence of violence and conflict? It seems like to support his own position, Harris has to resort to a willfully oversimplified notion of human nature. People fight all the time. They do it for power, resources, ego, religion, or whatever. More often than not, it's probably some combination of all those. As an argument for why people should disbelieve in God, I don't think the fact that people are disposed to come into conflict with each other carries a great deal of weight.