I do not see "women can't code" anywhere in that memo. I see "women are less likely to take interest in coding (for biological reasons possibly)" which is completely different thing.
Granted. To this point, I'd say this: more than half the internet read his memo and understood "women can't code". At some point, it doesn't matter if that's not what he said. The impact of his work is how people perceive it, not what he meant. This isn't really fair, but it is so.
One thing that's obvious: the author clearly did engage in broad generalizations about half of the population; your quoted sentence is proof. He hedged these statements with others, saying that of course it always depends on the individual. But when a person makes broad generalizations about populations, they're always playing with emotional fire. And once people get bothered and emotional, all of the rational hedging in the world won't save your job.
One thing that's obvious: the author clearly did engage in broad generalizations about half of the population; your quoted sentence is proof. He hedged these statements with others, saying that of course it always depends on the individual. But when a person makes broad generalizations about populations, they're always playing with emotional fire. And once people get bothered and emotional, all of the rational hedging in the world won't save your job.