Earlier quoted context omitted.
How does your "gender-based difference in interests" explain the last thirty years of (proportional) decline in women majoring in CS?
30 years is plenty of time for the interests of a demographic to shift.
That's why trends are just not that useful in convincing people, especially people whose job it is to be logical week in, week out, and to generalize aggressively.
So these people look at you say "the trend is down" and then see you make that a primary reason for action, and then they immediately go and look at all the trends, and then notice the participation trend is down even more in the humanities among women (also having peaked in the early 80s, which makes me wonder what changed then), but regardless, they also see that the downward participation trend in the humanities is not a reason for action there, and wonder WTF? It seems like special pleading, because it is.
There are good reasons for promoting CS to females today (as ever), but "the numbers show a downward trend in CS" is not one of them.