Have to disagree with post that turn coordinator is inferior to turn rate indicator. By canting gyro axis to incorporate roll sensing as well as yaw sensing, you get a much better response to small roll inputs. You are also much less likely to overcorrect when rolling to wings-level after an accidental steep bank. Everythings works better and more intuitively. Also, precise turn coordination (rudder use) becomes less important as the instrument becomes much less sensitive to adverse yaw.
There's a reason that pilots used to be taught to fly the needle primarily with the rudder, and the ball with the stick. It's a clunky and imperfect system, (and completely dependent on some degree of yaw-roll coupling via dihedral etc), but at least you don't see a temporary wrong-way swing of the needle due to making a strong roll input with the stick and failing to give enough rudder to eliminate adverse yaw. Not nearly so much of an issue with the "turn coordinator".
Of course most modern airplanes don't have much adverse yaw, but the principal still remains...
There is some danger of an un-practiced pilot reading the turn coordinator backwards, like an artificial horizon. This could be fixed by going back to the old style of display (needle an ball) while keeping the innards of the modern instrument. Or, by practice. Or, by painting a horozon depiction on the background disk and making it roll left and right, while keeping the airplane in front stationary. Like an artificial horizon.
I think pilots should get more chances to practice partial-panel flying in actual IMC...