I agree, and this happens on an even not-slow connection. I would be interested in seeing the testing studies that made them go with this change. (I’m sure they did some.)