A botanist friend of mine had the insight that you can’t outcompete nature on things nature cares about, only on things it doesn’t care about. So, you’re not going to get anywhere researching better photosynthesis because if it were easy, nature would have figured that out a couple of million years ago. But you can make a larger fruit, an easier to store grain, etc. because nature isn’t solving those problems. And nature has a lot of trouble solving coordination problems, like how do group selection so you get a crop that grows well together without just selecting for the plants with a 1 foot longer tap root that steals all the water from their neighbors. So that has been his research area.

Open source is similar to nature in evolving largely organically (although people do make conscious choices to intervene in it, so it’s not a perfect analogy). You’re not going to beat open source at certain tasks, but you can definitely beat it in coordination. Of course, once an organization becomes big enough, it gets its own internal ecology and you get principle-agent problems, and the whole problem just recurs at a different scale.