Someone correct me if I’m wrong, but didn’t we get into this mess when too many people in 2003 and 2004 extrapolated recent trends and then made prediction based on those extrapolations?

Exactly, hence the value of looking at the recent trends within the context of the long term trends rather than using ant point in a single cycle as the reference point.