You can learn a lot about a candidate, particularly those applying to senior positions, by looking at how they work under uncertainty and time/resource constrai

You can learn a lot about a candidate, particularly those applying to senior positions, by looking at how they work under uncertainty and time/resource constraints, because ultimately that's how real life is. I have seen software engineers getting lost trying to solve a trivial problem because they like to always write code that scales well or showcase their knowledge of software design. Take home projects are great to learn about a candidate's decision making process given the use case and constraints of the final solution. IMO that would be hard to assess by looking at a repository. As far as having an entire team of top performers, IMO most companies don't need that. Where I work most of the stuff we do never gets released as a final product, we may work on a new feature for six months and then gets killed, because the business requirement changed, because we partnered with another company that already provides that functionality or because our VP changed his vision. Having a team of just top performers would not change that fact, but you will still have to pay top salaries and get into bidding wars with other companies trying to poach them. We do have a few way above average coders leading the way, but most of our team is made of good coders, not stars. So far that's working great for us.