Though it does seem to show decoding is slightly faster with LZFSE unlike your parent's local benchmark. It's not clear to me that either is always superior on those charts.