Well that is actually not true, scoring from crosses depend on both crosser and the guy who receives it at the end.
I'm not even a Perisic supporter, but here we go.
Just because the receiver has managed the ball badly on the receiving end side, doesn't mean the cross itself is poor.
The same reasoning can be used to address Politano's lack of throughput with his crosses. Although through observation, he underhits his crosses so often that it doesn't even reach the target.
Basically we are talking about the wrong type of data, using it wrongly to support the wrong argument here. Basic assist numbers just don't show crossing throughput, it shows the crossing throughput in conjunction with the finishing throughput, but again this is receiver dependent.
The argument here is Perisic can still cross, but he is unable to move the ball up field, evading opponent to get to that cross. This problem also affects the assist number, which is why the number itself is irrelevant to the quality of his crosses.