That "notable criticism" is terrible.
His whole argument is that because this algorithm does not outperform sklearn's HistGradientBoostingClassifier the paper is useless and a waste of his "very precious time"
But the authors claim "Hopular surpasses Gradient Boosting (e.g. XGBoost), Random Forests, and SVMs on tabular data." So if a simple sklearn model works better, it's worth knowing it might not be all it's cracked up to be before deciding to spend hours understanding another paper.
The fact that SVM was the #2 algo in these results, and that linear methods beat out all the other methods means there's something going awry.