His memo absolutely fails to mention the biases women deal with on a daily bases. It fails to mention how unjust such things can be and how they must make his female colleagues feel. And that such things are well supported by research.

That failure makes his audience significantly less receptive, and puts them on the immediate defense. He should spend some serious time reading "Difficult Conversations" and other books of it's ilk. As much as he wants to avoid feelings, feelings always matter. Hence the justified accusation of his memo being "tonedef".

And you are not considering the context in which it was written, or are missing key statements like:

  > For the rest of this document, I’ll concentrate on the
  extreme stance that all differences in outcome are due to 
  differential treatment and the authoritarian element 
  that’s required to actually discriminate to create equal 
  representation.

For him it is clearly table stakes that women are discriminated against. He doesn't feel the need to argue it because he, and the Google culture which he is addressing, finds it so imminently obvious.

Something he states. Multiple times. Throughout the document.

  > Sure, he says that innate traits are only part of the cause. But the rest of his essay implies they are the primary cause.

The weight [of the research on the matter](http://emilkirkegaard.dk/en/wp-content/uploads/Men-and-thing...) (a significant amount of it) indicates that interests are a primary driver in the discrepancy:

  Therefore, from the inter-individual perspective,the 
  individuals who pursue STEM careers are more likely to 
  be male than female. For example, assuming that 
  individuals within the highest 25% of a population 
  interest distribution are likely to make occupational 
  choices consistent with an interest type, the number of 
  women entering the engineering occupation, then, is only 
  19.5% of the number of men entering the field. This 
  percentage is very similar to the actual female–male 
  ratio of individuals employed in engineering. In science 
  and mathematics interest distributions, the female–male 
  ratios in the upper 25% asymptote are 0.60 and 0.64, 
  respectively. However, the actual female–male ratio of 
  individuals employed in the field of physical sciences 
  is only about 0.40 and, in mathematics, it is about 
  0.45. This discrepancy between interest data and real 
  employment composition indicates that there may be 
  reasons other than sex differences in interests that can 
  account for gender disparity in science and mathematics.

A few observations though:

1) Google tends to pull a lot from Science and Math, not just engineering, in the "tech world". Dalmore is one such individual. This indicates the ratio should be much more skewed to an even makeup.

2) Even without that Google is still pretty "male" compared to what the research indicates.

3) Research on interests is largely still post social, so we don't really have a good feeling for the biological vs social components that drive the interest rates we see today. We know some pre-social effects exist, but we don't really have a good grasp on the magnitude of such effects. His memo fails to make this differentiation.