my real concern is that many authors use chill subs to have access to these for free, and removing them for the vast majority of journals seems like a questionable way to deal with big error bars