As far as EQ'ing orchestral stuff, it's really "to taste". Basically, increasing or decreasing parts can have an effect on the tone of the ensemble. More highs can add more airy or bite, more lows can (if careful about mud), add more warmth.
Certain sample libraries are known for having certain characteristics. EWQL last I knew was known as having too much mids. Reducing this can help fight mud and add clarity, but remove some warmth. It's all about the coloration you are going for and your personal tastes.
The other use is to get instruments out of the way for others, like trying to mix cellos and horns when they're in the same register. When dealing with a real orchestra, you can't turn a knob to fix that (but you can use mutes, dynamics, and 8va/vb to fix it as people have done for centuries), so I don't know if that's really most applicable to orchestral music as much as electronic and pop styles. I would assume EQ for orchestra would be a lot more subdued than mixing a jazz or pop track where you have different EQ for each drum mic and compression and all that jazz.
When dealing with low-quality/free libraries or samples, it's almost required to EQ the tracks to improve the tone.