OKC has the best defense in the league. Cavs the best offense.

Link to tweet

It's not clear how the "net" is derived.
And, there are some fairly pronounced differences between team with the same record. (Spurs compared to Warriors, or Lakers to Bucks). The Kings rate higher than the Mavs while Dallas has a better record.
It feels like someone digging into stats without understanding context.
The teams at the extremes seem tautological.
The chart seems like the answer to a question nobody asked.

The defense/offense numbers are rank orders from best to last.

So the chart is showing where a teams point differential is coming from, more from defense or offense or a mix of both.

For example, CLE is +11.5 and OKC is +12. Very close but got their in very different ways. CLE is 123/111 (off/defense) and OKC is 116/104. CLE 123 ppg is the best in the league, and OKCs opponent ppg is essentially tied with ORL as best. Thing is that ORL averages 105 ppg.

Point differentials correlate very closely to win loss records. In the east all the top six teams have positive diffs, starting with CLE at number one. Same is true in the West. As the season progresses, the Law of large numbers works it's magic and the correlation moves closer to 1.

There have been teams where this correlation is weak, GS a few years back used to blow out teams fairly frequently, but also lost a lot of close games. So the diff was not very accurate because it was skewed by the blowouts.

That actually reinforces my belief that thus is a weak metric.