Oh it's really easy to extrapolate stuff. But perhaps they shouldn't. A lot of papers I've seen recently have wild romantic extrapolations based on cherry picked correlations.

Some of the social sciences are terrible at this. A former partner was a researcher in one of what I now consider to be less respectable fields and she would come up with a feely conclusion and fit the data to it and publish it. Wanted me to co-author one with her and do the statistical analysis. Told her I don't want to be on Retraction Watch.