A huge factor with academics, and with anyone in high-skill professions (critically those where your boss does not know how to do your job) is that in a lot of job situations, one has to perform their actual job functions, and then, separately, one has to perform (literally) a PR show for their managers, putting on an outward display of competence. Managers can measure competence up to the point of obvious failure, but once you get above that it gets FAR harder to tell if someone's good, and the whole affair becomes susceptible to "con men". You see this with professors, you see this with engineers, with doctors, sometimes with lawyers, etc, etc.
So even if you're acting in good faith, you need to put on a theater act of being skilled at your job.
Here's the thing: It's a million times easier if all of your teammates are supporting you. If you're in a cooperative relationship with your teammates, then your teammates can "allow you to admit your mistakes". Quietly amongst your peers, in a way that doesn't hurt your chances of promotion. But the important thing is the mistakes get exposed AND FIXED. Also - the confidence intervals get exposed, and guarded against.
In an adversarial relationship (which you see in more zero-sum games like professors fighting over tenure), people have to hide their incompetence, and the horrible, dangerous thing is:
—they get away with it—
… because the time-scales are long, and the confidence intervals are high enough that it's a reasonably safe bet. They might have a 1-in-5 chance of having it blow up on them, but if they talk a big game, exude confidence, push a risky bet, and have it work out, they could easily become a team lead, and basically seal their career. Life is short and lots of people take those risks.
Being cooperative simply boosts competence; not merely as a team, but as an individual. Since they don't need to pretend they're always making the right choices, people don't have to double-down on "inferior choices" to avoid pretending they're wrong. They can just shamelessly change their minds and pivot to being more correct.
Any sources for this? Should we assume the '96 Bulls were "best buddies"? Are Amazon and Apple the "most tightly knit companies on Earth", thus reaching $1T valuation? This sounds like an "I assume this to be true by' common sense', therefore it's persuasive" type of claim.
The problem is, competence is important. Nice people who design airplanes that are "mostly right, but fall out of the sky slightly more than average" have real costs versus a team that's "a little cold to each other, but meshes well to create a high quality product."
Which matters more probably ultimately depends on the "mission criticality" / ramifications of failure of the product. But assuming that two Advanced Beginners will elevate each other to competence also ignores that they can create a "blind leading the blind" style of effect.