Come to think of it, I'm sure I can recall gliders which required positive out-stick to stop their tendancy to roll into the turn, however, that doesn't explan why all the short winged noisemakers I've flown *do* seem to require a little in-turn stick... and I do know what a co-ordinated turn looks like
I have a theory but you might not like it.
You don't need to a balance a turn to make an aircraft yaw. You can leave an aircraft slipping out of the turn and rely on the directional stability to yaw it round. That's what happens if you don't apply the rudder necessary to balance an already-established turn -- note I'm not talking about the much more significant rudder input often required rolling into/out-of the turn because of adverse yaw.
The amount of slip for a given turn-rate or bank angle depends on the airspeed and the coefficient of directional stability. In a glider, trying to drag it around a turn in a slip is an ugly, noticeable process, and you probably use rudder carefully to balance it. In a short-winged noisemaker, it's much more difficult to detect the slip, and less rudder is generally required to balance it, so little in fact that we often don't use it. And the balance ball is not very sensitive. So we actually tend to fly them around with a little out-of-turn slip, which produces an out-of-turn rolling moment, which at low bank angles outweighs into-turn roll of the differential wing speeds. I reckon if you used a slip-string to balance the turns, you'd find a tendency for the aircraft to roll into the turn.
As you implied Mark, one good test is worth a thousand opinions. I'd encourage everyone to look at the aileron input required (i.e. control-wheel/stick position) next time they try a steep turn. I'd be surprised if you find that in-turn aileron is required to maintain the bank angle.