If a kid does not get cookies at home it can shoot all its classmates. Lots of media coverage, the kid will get "the attention of an incredible number of people very quickly".
For me, publishing code under a FOSS-license means giving back to the community. Anyone who then decides to cause collateral damage in that community was never in for the community in first place. Sorry, that sucks. It causes extra work to developers who rely on your code. They have to spend time to fix a non-issue. Time they could spend with friends and families. If I was a js-coder, the author would be blacklisted for life. If I was an employer, the author would not get permission to publish code created during work hours under a FOSS-license without the company having a private repo. Such a reaction actually costs real money to lots of people. And it is a great disservice to the FOSS-community because it sets a precedent.
Really, comparing this to mass murder, well done. It's more like he had a bunch of toys, someone stamped on one of them and he took the rest home with him whilst the other kids are still playing with them.
These are his repositories which he created, he is welcome to remove the code, and others are more than welcome to rehost (as they require).
If you are depending on npm (or any other build tool such as composer, whatever) or on Github, you really need to have a backup plan when the shit hits the fan.
The mass murder example is a counter point to the (implied) argument in the grandparent that "[getting] the attention of an incredible number of people very quickly" is the same thing as an effective protest.
It is possible to get lots of attention for your action without that attention translating to support for you or your cause.
The parent continues to discuss how they believe there was damage above and beyond "wasting the time of a bunch of build cops" and explicitly states what they think the damage would be.
Perhaps choosing a different example would have been more tasteful, and may have avoided this side discussion and being flagged to oblivion, but it did not ever claim the two events were equivalent.
It's possible that the mass murder example, while not directly compared with the original action, is implied equivalent by the mere juxtaposition of the two but I don't think that was the parent's intent.
If someone argues that some $noble_end is served by $means_under_debate, as a justification for the means, that argument can be refuted by giving a circumstance where the same $noble_end is served by some $extreme_means where both parties can agree that $extreme_means is never justifiable.
Then once we've agreed that the form of that argument doesn't work, it is reduced to 'the ends justify sufficiently noble means', we can move on to discuss whether the $means_under_debate are sufficiently noble, setting aside the fact they they lead to $noble_end.
As far as I can tell his code is still up on github. He removed it from a poorly implemented distribution channel on which many people happen to unwisely depend. I'm amused at how much people are concerned about this breaking apparently production code. It will serve as a good lesson to many inexperienced and lazy engineers, I suspect.