As an introduction, the following are some publications covering various aspects of the Google diversity controversy:
I liked the article by Megan McArdle. However, I was asked to express my opinion about the last article, by Cynthia Lee. Cynthia's point of view appears to be representative of the uproar, and is phrased in a cogent and non-extreme manner.

Broadly, my views are as follows:
  • Throughout centuries, society has historically done an extremely poor job accommodating minorities of all types. There was a cultural and social straitjacket that worked only for people who were cis, straight, not disabled, the right color, and followers of the dominant religious and political doctrines.

  • To make a short detour into communist (socialist) countries - this was no better. The main difference was a change of dominant religious and political doctrine (atheist vs. Christian or Muslim, common property vs. private property), and change of role expected of women (no more traditional housewife, now everyone worked in factory).

  • In recent decades in Western countries, there have been great strides in the accommodation of minorities. This has been an enormously good thing. Homosexuals, transsexuals, people with disabilities, people of color, and career-oriented women have all made great strides toward acceptance.

  • Since the great strides in acceptance are all relatively recent (remember that Jim Crow laws in the US persisted until the 1960s!), there is still a significant cultural backlash against these improvements. This backlash is to a great extent bigoted, and negative. There remain large segments of the population who hate homosexuals, transsexuals, successful women, and black people, and want to deprive them of acceptance. This is traumatizing and awful.

  • Because of the existence and awfulness of this backlash, it is necessary to stand and fight against it.

  • Unfortunately, the people who stand and fight seem to forget they are fighting for equal rights of minorities. Instead, it becomes a fight to marginalize the majority, to the point that it should be invisible. The mere mention that there is a majority at all – the mere non-aggressive presence of it – becomes offensive.

  • In software engineering, women are currently a minority. This is not necessarily due to bias. Women are also a minority in e.g. the US prison population (around 7%), and in dangerous occupations like logging (3.2%). Although some – many? – believe this is due to pervasive sexism, this is not necessarily the most complete explanation.

  • Cynthia Lee is among those who appear to believe there's a gender issue until women are represented at perhaps up to 50% in her field. This may be ill-motivated, and may not take into account women with different interests. Just because Lee is at the top of her field at Stanford - and she should be accepted, and her qualifications should be respected as such; and other women should have the same opportunity - that doesn't mean that the trajectory of every woman should mirror a man's.

About 18% of computer science students in the US are women. But outside of R – a language mainly used for statistics – women are much better represented in web development languages, such as JavaScript or Ruby, than they are in systems languages, such as C++ or Erlang.

Why don't women pick up C or C++ in greater numbers? Does this have to do with that these languages rely heavily on pointers, which is helped by spatial reasoning? There are documented differences between genders in spatial abilities, and it's one reason some women quit engineering. But women, for example, are much better at distinguishing colors.

Spatial ability can be improved with training. But if these differences are overcome – what if the gender gap persists, simply because there's a difference in interest?

Is it permissible for people to differ in interests?

And if yes – why do we expect these differences to average out to nothing?