I have rarely seen a balanced pair of opposing cabals. One side usually wins out, at least on individual articles.

> watching like a hawk for their opponents to make errors they can correct

This is a very idealistic view of how these battles are actually fought. Just settling for correcting the opposing side's errors would be very magnanimous. In reality, cabals usually try to get the opposing cabal's editors banned. The types of errors they look out for are violations of editing rules. Usually, both cabals are in violation of the rules, so it's really a question of which side is more effective at gaming the system. It's very common for both sides to try to bait their opponents into violating the 3-revert rule, for example.

Once one side wins, they can do pretty much anything with the article. Most articles about controversial subjects end up under the control of one side or another.

There are a lot of cases where a small minority is heavily invested in a topic at the individual level, while the majority (who have an opposing view) are not. The small minority has the incentive, the energy, and the time to keep fighting for their narrative, while nobody in the majority makes the same effort as none of them individually have as much skin in the game. This dynamic plays out all the time on Wikipedia.