In other words, you can't predict the team's Net Rtg for next year on last year's Net Rtg. Therefore, it's a worthless exercise to use it to predict next year's wins and losses.
Correct?
The first bit is more or less what I mean. My point is that Net Rtg (or very similar stats, like Margin of Victory) does an amazing job of correlating with record, it just doesn't do nearly so good at predicting the next years Net Rtg.
For instance, if you calculate the correlation between a team's MOV (easier for me to grab than Net Rtg, but same idea) and its record over just the last year, the correlation coefficient is 0.96, which is huge. You can explain 96% of a team's record simply by looking at its point margin. But if you compare teams' '17-'18 MOV to their '16-'17 MOV, the correlation coefficient is only 0.62. Yes, that's not at all bad, but that suggests that you can't capture over a third of the variability in MOV (and therefore about the same amount of variability in record, given the tight correlation between record and MOV) by just comparing to the previous season.
There are reasons for this, not the least of which are player movement, changes in coaching staff, addition of draft picks, player development, player aging, injuries and recovery from injuries, etc.
Just to belabor the point, the fundamental difference between the two comparisons is that in the first case (Net Rtg vs record), you are taking two different measures of what has already happened and seeing how well they agree, but there is no prediction. Both measures are taken on the same results from the past. That the agree so closely suggests that they are closely related, as we should expect point margins and record to be, as wins and losses come directly from point margins. In the second case (Net Rtg vs. Net Rtg subsequent year), you're trying to use one measure to predict an outcome that hasn't happened yet, and there's a lot more uncertainty in that.
I don't think I would call anything here a "worthless exercise". But I think we need to have an idea of how much value the exercise has, and what the pitfalls are.