I don`t Eq to the busiest part.
I EQ to fit stuff together.
If a sound needs extreme EQ to fit it in the mix, then it is a bad sound choice, and I take it out.
EQ is as much choice of sound as it is an actual filtering process.
Some sounds are a little larger and require more room, but if you are doing things properly the subtractive EQ you are using shouldn`t be so destructive that sounds on their own don`t work.
It sounds like your cubase instructor was either, overdoing his EQ, was working with a bad singer, or a badly recorded singer, as he was doing very destructive EQ.

When it comes to mixing bands it`s a different story in some ways, the human voice is pretty complex tonally, and if you screw with it too much, it is more noticable.

I just got a hold of the entire multi track recordings from the Latest NIN album, and I was checking out the vocals, as I am going to do some remixes. The vocals in the multitracks are all post EQ, and despite his music being very heavy on large, complex electronic sounds, the vocals sound fine solo`d, as the engineer did a proepr job.

It sounds like what you are talking about is if you overdo the subtractive process.
I try not to make any adjustments over 2-3db in EQ, so if a sound does drop out, it simply drops out and doesn`t leave a huge gap.
It`s like if there are 2 of you in a room talking.
If one leaves, the room will become quiter, but the whole dynamics of the room shouldn`t suddenly change.