Correlation is not always causation.
What benefit does Oxbridge have, as elite institutions chasing the most academically gifted candidates, to go out of their way to discriminate against the female applicants? If anything, they are interested in making the numbers as even as possible.
Other factors are the cause here, and it's impossible to truly know which. Perhaps the MAT/TMUA scoring of male applicants in that particular year was stronger than the female applicants. Perhaps it is a low adoption of advanced programming languages with females, leading to lower scores in the interviews and lower demonstrated interest in the subject.
By the way, the figures you've cited show a higher shortlist to offer conversion rate for females (41% to 45%), and the highest conversion rate is with Other/Prefer not to say at 52.8%. Could I not argue that Oxford is actually more favorable to get into if you go gender-blind?
Try not to fuss over admissions statistics. I saw someone post this week about comparing Oxbridge acceptance rates while failing to account that their intakes are completely different. There is more than just numbers.