However, there have been a number of criticisms to the paper about data coding, methodology, and interpretation/motivated reasoning. Recently, the author has also acknowledged that governments have adapted since 2006: her data says nonviolent resistance is still more effective, but that the success rate has cratered.
I tried to find some of the criticisms, but it's recent, so all AI generated, so I won't link before reading, but the criticisms I'm aware of I think are robust. All this to say, the idea that nonviolent resistance is more effective is comfortable and feels better, and there is a paper that says so, but your examples and those of a child comment are not about resistance to armed colonial oppression and it's not enough to form a blanket opinion about violent resistance.