Analysis
It's simple. In 1997 (the year before Dirk was drafted), around 84% of shots were twos (threes only made up around 16% of all FGA). The majority of those (58%) were within three feet. A center inside could contest up to 48.7% of shots by hanging out in or near the paint.
In 2018, only 68.5% of shots were twos (31.5% of shots taken were threes). A reduction in the importance of the mid-range jumper meant that a higher percentage of twos were within three feet (60% of twos), but that only made 41.1% of FGA that were in the paint and thus easily contested by a traditional center.
In addition to threes diminishing the number of shots in the paint that a center could contest, the increasing number of threes also means that more rebounds are ending up further from the basket. Those rebounds depend on high-mobility guards and forwards to rebound, further diminishing the value of the big, lumbering center.
TLDR version:
Since 1997, the number of shots inside that can be contested by a center has shrunk by around 8%
Since 1997 the number of rebounds available inside has shrunk (although there isn't a good stat to track distance from hoop when rebound is secured)
since 1997 the mid-range game has taken a huge hit. Teams are increasingly scoring either inside or outside the three point line.
Disclaimer
Close twos= I added up all shots classified as "dunks", "layups", and "close"
Midrange = I took the total FGA and subtracted shots from behind the arc as well as those classified as "close twos"
My analysis does not account for FTs, since it's impossible to determine where inside the arc the foul took place or even to find an aggregate of where the FTs came from, particularly going back more than five years. We know from experience that players are fouled attempting shots inside the 3pt line more than outside it, but it's hard to do a statistical analysis.
Sources
82games.com
basketballreference.com
nba.com
Graph linked from businessinsider