Below is the explanation based on the code test instructions I originally received.

From the instructions:

Write an application that uses Twitter’s API to find the top 5 most famous people to tweet about “PlayXFL” or “XFL”. The application should be runnable by us.

Methodology

• Definition of "famous people tweeting about PlayXFL or XFL". I restricted my search to verified accounts (to try to gauge "fame" versus simply "large number of followers"), and then sorted by number of followers from there. I excluded the XFL accounts themselves, to gauge people tweeting about XFL vs. from XFL.
• "The application should be runnable by us". Initially, I wrote the code as seen in dist/main.js, using async/await and the Fetch API and vanilla Javascript. However, this would only run in modern browsers. Trying not to assume the user has a modern browser (while taking advantage of modern Javascript), and using Babelify on the backend to create a bundle for older browsers, the app is selectively loading the bundle based on elementary feature detection. Also relying on basic CSS.
• Security: For the purposes of this challenge, the .env file containing the Twitter API token has been committed to version control. In a production environment, however, this would almost certainly not be the case.
• Accessibility: Included ALT tags on images for screen readers as well as maintained heading order.
• Rate Limiting: I excluded rate limiting for the purposes of this challenge, however, in a production environment, users would be returned the same content if they were to endanger the app's rate limit for Twitter's API.

Running

git clone git@github.com:ylove/boomtest.git
cd boomtest
yarn install
yarn build
yarn start

This will create a server running at http://localhost:3001.