Eh. It depends on the rule set - go has many sets of rules, all very similar and developed in different countries and for different tournaments. In most of them, KataGo not only thinks it's won the example game, it HAS won the example game. I'm disappointed with arstechnica for not pointing this out (even if it's true, as I expect it is, that this case study has implications for adversarial strategies in other contexts).