Good lord. 24 player sample and the chart even has "0.9" listed twice on the y-axis.
There might be something there, but this is beyond crap proof of it. Move the needle one inch in height (I'm pretty sure 6'2" is still short in the NBA), or one player on either side and I guarantee it would invalidate the whole premise. Not to mention correlation vs causation- how many of these guys suffered major injuries, what style of play did they have, how did they fare in measures other than WARP...).
It's worth noting that one way to read this chart is: "short PGs are practically identical at age 28 as they are at age 31", seeing as the line goes flat there.
It's basically a lazy quip about some numbers lying around.