Received a PM asking me to reply, specifically as to the "would" vs "will" stuff.

I agree: E on this one. The author's indicating a state of not-quite-sureness about this whole warming thing. He expresses this in two forms:

1) whether something will happen

2) what impact if would have if it did happen

Both constructions (notice the italics) indicate some kind of future uncertainty - either "whether... will" or "would."

Both of the things the sentence is discussing should properly indicate future uncertainty - that's the point of the sentence, right? So you could use "would" twice" or "whether... will" twice, but they chose to do one of each in the correct answer here.

We can eliminate A and D for the above reasons.

Next, the scientists are debating both of these issues, not one or the other, so we want "and," not "or." Eliminate C (and D, again).

Because the scientists are concerned about both of these issues, those issues should be presented in parallel.

B uses "whether" and "the impact" to introduce the two things.
E uses "whether" and "what" to introduce the two things.
E wins.

Re: the divided on / divided over thing. I think the original idiom was "divided over" but "divided on" has become quite pervasive in the language, enough so that I would have expected even GMAC to accept it. I wish GMATPrep had explanations so that we could read their rationale.

I'm least enthusiastic about that last one. I've seen better questions, overall. (And, annoyingly, I think this one is GMATPrep...)