Well, noncognivism is basically Hume's guillotine (You can't derive an ought from an is, basically) in fancier terms, from my oh so shallow knowledge of both.
Valid reply, I'm sure I'll see all those points come up, but I think they're also all pretty debatable. Education and 'we need to assume certain things about moral truth' are probably the only arguments that will actually stick to the case. Also, 'we inherently value things' becomes 'we value things due to biological processes' = evolutionary morality = naturalistic fallacy. Just because you can tell me how I'm hardwired to act doesn't mean you can tell me how I should act.
I think I'd probably try to absorb your 4th point, since I'm valuing Moral Autonomy and arguing that with lack of definitive moral truth we have no grounds to impose an obligation on an individual's right to decide for themselves right and wrong. As for the hypothetical card, if you want to ride the skepticism train too hard, nothing is said to exist for certain, so I hardly see how that's going to make much of a point.
Actually, this would probably help the morality isn't real case. Because Morality isn't grounded in facts. It's grounded in a lot of assumptions about what constitutes 'right' and 'wrong' - furthermore, since the resolution is makes this a debate of abstract moral concepts, saying that 'theories not grounded in observable facts' are off bounds seems to be both illogical and against education.