Although a side point to the main point of the article, I draw your attention to the tables in about the middle of the article.
fWAR produced across MLB by various classes of players:
Pre-ARB: ~40% of total fWAR
ARB: ~30% of total fWAR
Players who reached FA: ~30% of total fWAR
Total amount of payroll across MLB tied up in various classes of players:
Pre-ARB: ~15% of total payroll
ARB: ~25% of total payroll
Players who reached FA: ~60% of total payroll
One way of looking at that would be: $ spent on pre-ARB players are about 5 times more effective than $ spent on FAs, and $ spent on ARB players are about 2.5 times more effective than $ spent on FAs in terms of generating fWAR.
If you were wondering why having a lot of young, talented ARB and pre-ARB players was important.