Thank you all so much for your detailed responses. I will certainly take them all to heart when forward.

Your concerns are completely valid, and has you have no came to discover, you don't 'learn code'. Imitation is the best form of flattery, but it will get you no where. If you do not understand the underlying principles( theories, postulate ), behind what you are doing, then saying you learned it from code is just a coverup for copy-pasting. So how is your proficiency with:
-linear algrea
-lighting

-shading.
-..insert any other game development related atoms here.

If you can truly say you have a good grasp of these by just looking at other code, then I'd say go ahead and knock yourself out....If not you really need to evaluate what it means to learn something. Learning and understanding are 2 separate concepts, hopefully, you are not confusing the 2.

Is your goal to just learn or also to end up with something usable?

You mention that you want to create a small turn based rpg, but you also talk about writing a game engine. I'm actually working on a simple rpg too and I think an important part of the process is to decide what you are making. So are you writing a game or an engine? To ask bluntly, what is so special about this rpg that you need to use code written for professional game engines? Arguably the code from those engines might not actually be suited for a small game. In fact your own simple solutions might be a much better fit, they will be easier to understand and maintain. Take graphics for example: Do you really need the most efficient rendering pipeline? Unless the term "small rpg" means something different to you than me, I'm pretty sure you don't need that stuff. If you still want to figure out an optimal solution, I would say do that after you have written your own, the learning effect will be much better that way.