1) Simplifying concepts can be dangerous. I once solved an exercise where it was asked to calculate the volume of a cone, except that the rotating axis was the $y$ axis instead of $z$. The computations remained the same and I told them so, showed how all we needed was to switch the $x$ and $z$ coordinates to polar coordinates and the analysis proceeded the same as usual. However, I noticed they cling too hard to this: instead of a understanding of the situation, what I felt from their reactions was "Oh, so whenever I have this, I do this", and one more rote memorization. This may happen because multivariable calculus is an elementary course and most people taking it are far more interested in magical solutions rather than thought processes.
I've also noticed other departments do not value these as much as we do. I'm taking Electric Circuits Analysis this semester and I did not so as well as I knew on the test. Part of that was because I spent a lot of time explaining what I was doing and the processes behind my computations. I ended up with little time for checking the computations and ended up getting a couple wrong. When I first received the test I had a horrible grade, and was astonished to realize that all the text I had written was as good as trash: they didn't read any of it! The marks were solely for the numbers I had arrived. After I met with the professor he gave me some marks back, but it still leaves a bad taste. How can we teach people to think if we measure them solely by the results? The end does not justify the means.