I thought the same, isn’t a lot of this data stable and static. Why recrawl and continually index stuff that has low value if the corpus is already feature complete.