Years of studying what, though? Calculus, most of linear algebra, discrete math are all calculation based. Math papers are proof based. If you spend years studying proof based math (analysis, algebra, topology, and so on) then you'll be familiar enough with proof based math to understand it.
It's like saying "why isn't spending years studying spelling not sufficient to understand Ulysses?"
We spend years learning "spelling", because the mathematical equivalent (arithmetic, up through calculus, linear algebra, differential equations, discrete math, etc) is incredibly valuable on its own—and that that value doesn't really require familiarity with the proof-based foundations underpinning it. Learning real analysis is valuable if you're going to be a mathematician, but as an engineer it won't help you get more out of calculus (at least, not compared to the cost of learning it).
I believe you have adopted your opinion of what should go into the first set of brackets of yours.
An engineer is expected to allot too much time to math to turn out unfamiliar with it. I think we should spend less effort on practicing cleaning fish and more on learning to forage for fish.
It's like saying "why isn't spending years studying spelling not sufficient to understand Ulysses?"