while i agree with the spirit of the title, i find the specific conclusions dubious and the arguments presented unconvincing. if touch ui wasn't more convenien

while i agree with the spirit of the title, i find the specific conclusions dubious and the arguments presented unconvincing. if touch ui wasn't more convenient for _everybody_, it would remain a weird niche use case. so it's still building for the majority use case - it's just building the right thing, not a "faster horse", so to speak. the problem ive seen with data-driven design is overfocus on specific metrics, and the resulting tunnel vision leading to building "faster horses".