I agree with you that their answer seems poor, for the reasons you outline. They might be relying on "yesterday's weather". the XP principle that the best predictor of the weather for today is what the weather was like yesterday. This gives greater weight to more recent experience.
But generally, when using past velocity to predict current or future capacity for the team, we average the most recent three or four iterations. This has better predictive value than a single result, even if it is the most recent one.
Your reasoning is entirely correct. It is sad therefore that the answer sheet gives A as the correct answer. At the very least, you could have several "correct" answers for this, and good arguments for each one. Unfortunately, this is an example of a poorly designed question.