Chris Bertram at Crooked Timber has some interesting musings on conditional arguments.  Critical point:

Strawser claims that IF drones reduce civilian casualties compared to other means THEN the use of drones is justified (I’m simplifying). Philosophers will typically then say that the argument is merely conditional, and that therefore, if the antecedent is false then the conclusion doesn’t follow. Clearly that’s right. But does it get us off the hook in a world of propaganda, mass media, think tanks and the like? . . . .So, for example, I’ve heard it argued by philosophers that IF sweatshops improve opportunities for poor people in poor countries THEN they are on-balance justified: so people shouldn’t campaign against sweatshop labour. This then gets supplemented with “evidence” that the antecedent is true, but by this time the casual listener has been inclined by the rhetoric to accept the conclusion.

Here we have, I think, a major source for iron-manning: the conditional "arguments" are not really arguments at all.  They're conditional statements.  The real question, as Bertram correctly points out, is whether the claims are true.  As he notices, however, whether the claims are true is a secondary question (in the minds of some people) to conditional statement in question.  How those get evaluated is the more interesting question (to philosophers).  But it's often the wrong question.  And entertaining such arguments might often amount to a form of iron manning.

Here we have an example of this.  Yesterday Todd Akin, Republican Senate candidate from Missouri, remarked that in cases of "legitimate" rape, women cannot get pregnant.  Here's what he said:

"From what I understand from doctors, that's really rare," said Akin said of pregnancy caused by rape. "If it's a legitimate rape, the female body has ways to try to shut that whole thing down. But let's assume maybe that didn't work or something. I think there should be some punishment, but the punishment ought to be on the rapist."

I think it would be very hard to defend this remark, as it has no basis in scientific fact.  Sadly, if you treat the whole thing as conditional, suddenly it appears Akin is making an interesting point worth discussing among rational adults.  Here's Politico's David Catanese (tweeting):

"So perhaps some can agree that all rapes that are reported are not actually rapes? Or are we gonna really deny that for PC sake?" he said. "So looks like he meant to say — 'If a woman was REALLY raped, it's statistically less likely for her to get pregnant.' What's the science?"

Akin is saying something rather different.  He's saying that pregnancy is statistically less likely in cases of "legitimate" rape.  It's more likely when that rape is "illegitimate."  Catanese version has it that Akin is querying after some science.  As I think I've often repeated here (sorry), I think this is a kind of philosopher disease.  You're looking for the thing worth discussing, but in looking for it, you overlook or ignore the awful things before you.  So, yes, maybe there is a scientific question here we discuss, but that's not what Akin's point was.  In fairness to us, and oddly to him, we ought to represent his words and his intention correctly.  How else will he or we learn his "doctors" are wrong?

What's the harm?  Bertram poses an interesting question:

ADDENDUM: it would be an interesting psychological experiment (which, for all I know someone has done) to test whether people who are exposed to conditional arguments in the total absence of evidence for the truth of the antecedent become more inclined to believe the consequent, perhaps especially for cases where the antecedent is some morally dubious policy. So, for example, are people exposed to the conditional “IF increased inequality ends up making the poorest better off THEN increased inequality is justified” more likely to believe that increased inequality is justified, even when no evidence that increased inequality benefits the poorest is presented?

Anecdotal evidence says this is true.  If that's the case, then I think he might have an interesting point.