I am not nor do I want to be a physicist, though some of the stuff im interested in gets used a lot in mathematical physics i hear (moduli spaces, Floer homology, Fukaya categories).

Anyways, I am interested in hearing from the retards on this board as to the pros and cons of learning physics:
1) the "normal way" i.e. going through CM, EM, QM, statistical theory, GR, and QFT separately and building things up from easier to harder

as opposed to

2) starting from first principles like homogeneity and invariance and developing QFT, then deriving the fundamental equations of CM, EM, QM, etc. together, obviously at the moment GR would still be separate.

Clearly the first way is how its done so i assume its the better way to do so, maybe because it relies more on example/experiment and less math is needed at the outset.

However I think the second way may make it easier for people to understand the overarching themes in physics, especially if they have no need to "do" physics but want to learn about physics.

As a final note, in case some of you dont know what im talking about, here are two books that kind of illustrate what i mean by approach #2:

Principles Of Physics: From Quantum Field Theory To Classical Mechanics by Jun Ni
Physics from Symmetry by Jakob Schwichtenberg

What say /sci/ on this matter?