As someone from an academic background, it's a bit different than how academic labs are set up. My academic lab had a number of different projects in different research areas, ranging from human health to agriculture, but the unifying theme was big data analysis. The fact we all had this focal point meant we often had shared overlap with common tooling, which meant a lot of collaboration. Whether it was for alzheimers susceptibility or corn yields, you were working with tabular data.
Sure, we had people worked more on the bench and people who never set foot in the lab, but everyone made sure to know exactly how their data came into their hands and it's purpose, so if you were a statistician, you would learn everything about the corn sample you were given to analyze so you could make the correct considerations in your analysis. And if you were that wet lab person and wanted to present a figure that the statistician generated, you would learn everything about the test used, and all the assumptions made when choosing that method of analysis over others. Even in academia, this high level of collaborative interdisciplinary learning can be rare, but makes you a much better scientist who as a much better grasp on the wider project and your role to play.
I think a lot of startups operate with a mercenary mindset. Everyone is hired to play a discrete non-overlapping role, which tends to silo ideas. Central planning from upon high is also the norm, rather than collaborative discussion and solving problems from the bench up.
Depressingly, there are more and more big name academic labs that are adopting this startup oriented top down approach, with a head professor calling the shots and giving marching orders to a few sub research professors with their own postdocs, and grad students, and undergrads. I've known grad students and post docs in these labs who are outright denied to direct the research in their own projects, even if they have good ideas, simply because they didn't come from the top down. Pursuing your own ideas is the whole point of grad school and post doctoral training. On top of that, these labs siphon funding from more innovative and smaller groups by outputting higher numbers of ho hum papers, or affording expensive research with large, multi-institutional grants, both of which are heavily favored metrics in the grant proposal and tenure process.