I recently found Flax myself. I have no specific knowledge about the development behind it, however being an experienced game developer I have made several game engines from scratch in the past for various commercial game projects.
Very impressed with Flax so far, and I think the real magic behind such a feat is simply a dedicated, talented small and efficient team who are already familiar with existing engines and refine upon their experience with them. In the game companies I worked, we often put a tiny team in charge of a major part of the engine, as it is just so much faster than dealing with all the communication and overhead necessary in a large organization.
The typical Scrum framework many companies use these days tends to kill productivity, and being able to bypass it is super important to pull something like this off with few resources. But needs very talented and self motivated people, it is not for everyone. I would in fact say, it is not for most of todays developers.
I think it is no secret Unity (and to some extent Unreal) has been used as some kind of Blueprint here, and being able to learn from all the mistakes Unity did and take what works, remove the excess fat and improve the rest helps getting it right from the start. The time consuming part is often to have to experiment and try various solutions until your technical debt kills progress as all the small mistakes accumulates.
Also, in my experience, development is super rapid at first, then it slows down gradually as the product matures and the effort is more on maintenance and other aspects needs attention. One change in one end might do something unexpected in the other end as the engine becomes more complex. The first few months or even the first year is the fun part.
Anyway, just my two cents, the devs here might have a different story. 
. Maybe I am just paranoid?