Based on the benchmark, the "for in" loop appears to be 10% slower than the "for" loop. Does it make sense to replace all instances of "for in"? https://jsben.ch/kxBxv