I will not completely disagree about your +/- being useless, but i diaagree with you more than not.
This stat, and more important variations of It, will be seen more often. I also believe some of these +/- variation metrics will be more meaningful (or will continue to be tweaked until they are) and will more accurately show what is happening on the floor when a certain player is in the game. I can tell you from a players perspectve they "feel" what its like when their team is - 5 when theyre in the game.
I think your statement that theyre a completely useless metric is short sighted. Id be curious if CM views it that way. Didnt UR employ some Princeton stat guy a few years ago? Is he still around?
While more advanced plus/minus metrics may have more merit, the basic plus/minus stat is worthless. Consider the following:
Let's say there is player who contributes the exact average to his team, if he is in the game or out of the game his team plays exactly the same. Imagine, now, that in a game this team scores 2 points every minute like clockwork, and that their opponent scores 1 point every minute. A contrived and unlikely situation, I know.
If this player plays for 20 minutes then they will have a plus/minus of 0.
If they play for 10 minutes they will have a plus/minus of -10.
If they play for 30 minutes they will have a plus/minus of +10.
However, no matter how much they played, being in or out of the game actually had no effect on their team. For this player, if plus/minus was a useful metric it should be 0 in every scenario since the player being in or out of the game did not actually have any effect on their team. Instead, the different amounts of playing time result in very different plus/minus scores. The team was not better when the player was in, and was not better when the player was out, so why should different playing times result in different plus/minus metrics? This is one example of why it is a bad metric.
This problem may be solved by calculating a per possession plus/minus, which should result in a plus/minus of 0 in the three situations outlined above. That is a somewhat better metric. However, in practice even this adjustment does not make plus/minus that much more useful due to the huge variance in college basketball play. Teams do not score like clockwork, they score in spurts and droughts that are mostly outside of the control of a single player. The signal to noise ratio is very small, and the plus/minus metric is almost entirely noise, especially over just a few games. Kenpom discusses this here:
https://kenpom.com/blog/a-treatise-on-plusminus/