Like with all both filtering and EQ:ing, you must be very careful and never do a single decibel more than what's needed. With that said, I like the range between 8-10KHz! Many modern mics, like the Sennheiser MKH 40, DPA 2011, or Shoeps Colette-capsules are very straight and rich in range. That makes them sound great on big systems on a neutral sound-level, but most people will not listen to big systems with neutral level, so you'll have to make it work on lower levels as well without sounding sharp and muddy on high volume.
8KHz is a good such frequency to add a little boost of treble in a sound like that, it is produced in virtually any system there are, and is within range for all people not severely hearing impaired! You don't really get presence in that range, the presence lies between 2-4KHz approx, but it gives the voice a beautiful luster if not overdone.
To raise between 2-4KHz on the other hand tends to be quite annoying in the long run to listen to. It brings clarity, but it's the range we are the most susceptible to, so you risk giving the listeners ear-fatigue if used too much. And too much 3KHz also tends to mask important sibilants in human speech.
Bass below 150Hz approx and above 8KHz doesn't contain anything the voice needs for intelligibility at all, but it contains a lot of air and firmness. Removing too much of these tends to give a pretty dull and lack-luster sound. mixing sounds is pretty much like painting a picture, everything must have a purpose, but that purpose might very well be just to give wight and shine to the object, a painting with no highlights and shadows are pretty boring...
Also, while comparing to pictures, if a photo has too much purple in it you don't really lower the red and blue to level with the green as it would be too dark, you raise the greens. Vice verca, if it's tinted red, then red is what you lower. It's all a matter of opinion, but this is mine.