I can see why she chose not to use the raw date; perhaps the numbers in question are .0000001% vs. .0000002%. But that wouldn't have the impact of twice as many.
The author is using a fallacious arguement, implying that gun ownership causes murder by gun. While there may be a relationship between A (gun ownership by women) and B (being a woman killed by a husband/lover with a gun), it doesn't prove A causes B, any more than arbitrarily giving a student an "A" on a test suddenly makes him smarter.
I would argue that women at higher risk of domestic violence obtain guns more often than women who live in a stable home environment. Which makes this statistic meaningless to the author, as she is clearly writing about the use of guns in defending against a stranger.
My first thought as well.