What exactly do you expect to learn?
I don't think that watching other engines code is any helpful... at all(well I do It in order to see how "superior" my engine is
but this is entirely different thing).
I pretty much have no idea how an engine works. Everything I've done so far is mainly intuition.
I hate it when I code a lot then I realize I'm doing stuff the wrong way. I wish there was a good book on the subject.
I read main parts of Game Engine Architecture, mainly the "Rendering Engine" chapter. The author talks a lot about low-level stuff and not how a SceneManager should work at all.
I'm pretty desperate and open for any kind of information! Lots of motivation here
I don't think that "Game Engine Architecture" book is any good.
Currently I see 2 ways to learn "how to make engines"(trademark):
1 - Write a few small games(breakout, simple side scroller, "the impossible game", something minimal) from scratch. See what functionality you need and what API works best. After you've cleared your mind write a small library with that functionallity wriped up and call it "game rendering engine". This is want happend to me. It took few years, because I was 15 years old back then, and there were no references(or not much) at that time (Im 23 and I still write that "engine", this is my 3rd start from nothing).
2 - use a 3rd party reliable engine for a while, see what you like and what you don't and try to write a better version. This is just an idea from the top of my head, I don't know anyone who learned that way, and I'm not sure how effective will it be.
So if you really want to make a "rendering engine" my personal suggestion is going to be 1).