Those of us who grew up solving Newtonian mechanics, electrostatics, and calculus of 1-3 variables problems developed good intuition for understanding simple concepts. This helped me condition a problem, select a good coordinate system or frame of reference, and often provided a concrete example to more complex theories. For instance, it's impossible to picture a 100-dimensional gradient, but knowing a slope and a surface plane on a 3-D manifold makes it easy to imagine.
I'm getting to a point where the intuition is coming at a higher and higher cost as problems become more difficult. Differential forms are a good example, where the geometric interpretations are expensive to imagine and only really help in low dimensional space. I get the sense that quantum mechanics (or any "analogy" to it) becomes almost impossible to imagine.
For practitioners of physics: do you have to stop relying on geometric intuition, and just trust the mechanics/math? Or do you develop more mathematical intuition so visual analogies aren't necessary?