This advice might seem too simple, but the solution is to just start doing a ton of physics problems rather than trying to resolve any of these issues. Your heuristics for real world physics will probably change faster than you think, and it's very satisfying to feel like you really understand a common phenomenon in depth.

It does sound more like you want to study engineering from your latter paragraphs, but I'd argue a lot of that stuff is just industry knowledge you'd gather over time actually building things. Physics gives you the ability to do some back of the napkin maths to see if something seems to be the realm of feasibility though.