That's a vague and unconvincing chicken little argument for not making models using the wealth of open and for-purchase data that's available and understanding the limitations.

This is roughly where I've settled as well.

Economics likes to present the facade of a hard science with equations and rules - but macro economics simply doesn't work the same way as physics, and pretending that your equations are going to be a crystal ball has so far proved mostly bullshit.

Too many things depend on decisions made by actors acting under very complex and dynamic social constraints, rather than predictable rules.

The end result is that several decades after something happens - we're ok at examining the social/political influence of the time on macro economics, but that insight doesn't do much for us now because circumstances are changing at a pace that has rendered any intuition utterly obsolete.