What about in algorithm development or research? Could an average programmer invent quicksort, dynamic programming or information theory if transported back to the mid 20th century?
There is >10x productivity in scientists or mathematician using many kinds of metrics (papers, citations, awards etc)