Calculus and linear algebra both really only 'clicked' when applying them in my physics classes.
Physics is like the "word problems" we'd have in elementary sho, but for advanced math.
Physics' real-world origin prevents a lot of that sort of thing, and places emphasis on a practicality I find deeply appealing. Cauchy-Riemann equations are important, here's where they come from, here's what they do, bam done.
Let's forget about the pure/applied divide. That's not what I am getting at and I want to make my language very clear. Let's instead talk about "maths for its own sake" and "maths for x" (physics, computer science, what have you).
Calculus is a nasty kind of maths to study for its own sake. It's not intuitive, and it's shaped by motivations that aren't at all clear until you've learned its applications.
Stuff like abstract algebra or graph theory on the other hand are - to me - "quite nice". They're certainly much more intuitive than anything in calculus world.
So, basically I agree with you - in high school, maths (by which we really mean geometry and calculus) should be hand in hand with physics. I mean that is its reason for being in the curriculum, to prepare future engineers. People who want to study "real maths" can learn something easier to grasp, and maybe a bit more profound.
Superficial understanding :)
If you meant 'they don't REALLY understand it', neither do physicists. There are many layers that make up a computer, and most physicists only understand the very bottom ones, and they can hardly use a computer as a result.
Just because there are some parts you don't full grasp doesn't mean you don't know how the thing works. If that were true, all physicists would only have a 'superficial' knowledge, because even they don't completely understand how the universe works.
----------------------------------------------------
The first thing I learned in my Physics undergraduate studies was Maths. Calculus. Scalar fields and vector fields, their derivatives and integrals, and differential equations of all kinds and flavours. Non-cartesian coordinate systems, transformations between them, fourier transformations. And a nice toolbox for beating the not-so friendly integrals.
Even for classical mechanics, that's the foundation.
If you want to talk about Quantum Physics you also need linear algebra, in particular Hilbert spaces and operators. And distributions.
If you don't understand the Maths, you are left with talking about interpretations. Especially in Quantum Physics, that means you are left with basically nothing.
Without Maths, you are not doing Physics. You are doing Metaphysics.