yes because these benchmarks are not akin to the real world; rendering a large list of data? You'd use a virtualized list etc