> For what it's worth, many of today's finest, most important algorithms are from the 50s-70s, or maybe special adaptations of them.

This is only the case/sensible for algorithms that have not been practically improved.