To borrow a DevOps analogy, Gutenberg's website handles books as cattle, Planet eBook treats books a bit more like individual kittens insofar as they're hand-picked with (presumably) humans selecting the extracts and covers. The Gutenberg site has an early 2000s programmatically generated UI aesthetic that is similar to Internet Archive and the Wikipedia homepage. (At least in the case of Wikipedia, I've heard there's considerable community resistance to any significant UI changes - it's not just a budget constraint.)
A UX example. Search for "1984" on Gutenberg doesn't yield the book of that name. https://www.gutenberg.org/ebooks/search/?query=1984&submit_s.... Planet eBook site does: https://www.planetebook.com/?s=1984
To be clear, Gutenberg has been an incredible resource and gift to society. It's just that there's a need for custom front-ends to match different needs.