Your answer to part 1 is correct.
Your answer to part 2 is thoughtful, but not quite correct. In the absence of more information, the calculation you've done doesn't really add anything to the baseline information, and in particular does not show that their probability of having MMV is low.
Also thoughtfully done. A non-inferiority approach is specifically for when you are hoping to show that two things are similar. But it sounds like you think they're different. I would like you to sharpen that. Your hypothesis could just be: the difference is big compared to some standard. But I suspect that your hypothesis is really: sex A is bigger than sex B. If that's right, you can just say so. It's still good to have a standard for a size difference you think is important, so that you can compare it to your confidence intervals.
When you are talking about CIs “falling outside”, I think you mean “entirely outside”. Is that correct? Probably better to say so (or something else unambiguous, if you mean something else).
Grade: 2