One of the commonly-stated reasons for why discrimination persists in a field is that people making hiring decisions are biased. They might not know that they're biased, but they have some notions about the kind of people they want to see working for their company, and they disproportionately hire those people. This is quite an insidious form of discrimination because it's hard to prove, but it is possible to catch people out - for instance if they display a preference for talking to or dealing with male colleagues over female ones, or give more rewards, perks or encouragement to male colleagues. There doesn't need to be a "guys only" sign above the door for this to be a discriminatory environment.
Now, the original "Girls, girls, girls" post talks about how "super exciting" it is for KA to have hired twice as many female interns as male ones, how the women at KA network together, how there are exclusive events for female staff where female leadership is advocated. I know what Jessica was trying to say there, but damn if it didn't wind up looking somewhat like the stuff I described in my first paragraph. Now, I do not believe that KA has done anything wrong, but I don't think it's unreasonable for some people to question this.
This (as I post every time there's a sexism thread, I really should just get a blog or something) is all controversial because people are applying different ethical rules. Some people are applying deontological ethics, which says that there are rules about how to behave, and we should follow those rules (e.g. don't discriminate based on gender). They look at the "Girls, girls, girls" post and see things that look like they might break the rules against doing the stuff I described in my first paragraph. They're not actually looking at the "big picture" because they don't believe that, philosophically speaking, the ends can be used to justify the means. If bias is wrong, being biased in favour of a minority group is still wrong, in this view.
There's another more pragmatic view, which says that we should care mostly about outcomes, especially if no particularly great harms are being inflicted in order to achieve better outcomes. Although I generally lean more towards the rule-based view, in this case, I think KA is pretty clearly producing a good outcome and even if there is some bias[1] then it's not really the end of the world. But I wouldn't condemn or insult someone who thought differently.
[1] A key point about bias and discrimination is that people aren't (and often can't be!) aware of their own biases. If people from KA want to claim that they're totally without bias, that's going to make it hard to accuse other people of bias in the future. This is, sadly, just a Hard Problem and the various problems inherent in solving it are why we keep having lengthy discussions like this.