As Organic Marble notes, some single-engine stages do use RCS for roll control since the requirements for roll on ascent are less demanding than pitch and yaw.
The Falcon 9 first stage also uses RCS to control its attitude during the early phases of its reentry trajectory, where the grid fins are less effective.
It's possible in principle to use RCS for pitch/yaw steering on ascent, but gimbaling is more efficient. To steer a massive launcher, you need a fairly large amount of thrust, and thus fairly massive engines; they'd outweigh the gimbals and actuators. If they're fixed perpendicular to the longitudinal axis of the vehicle, then using them is somewhat wasteful: the force from the main engines and the RCS are the two legs of a right triangle, and the net force is the hypotenuse. With gimbals, the main engines just steer directly to provide the desired net force.
For a given deflection angle $\phi$ the total force needed from a longitudinal engine and a perpendicular engine is $\sin \phi + \cos \phi$ times the desired net force if the perpendicular engine is the same distance from the center of gravity as the gimbaled engine. For an average steering angle of 1º, this isn't too bad: only a 1.7% penalty. CG moves downward on the vehicle during flight, though, so it might be possible to do a little better by putting the perpendicular thruster at the top end of the stage; CG will still probably be a third of the way up at max Q, where you need the most steering authority, but you might be able to double the "lever arm" of the perpendicular force, amplifying torque, so the penalty might be under a percent. These lateral engines probably can't be as fuel-efficient as the mains, but we'll ignore that because 1º average deflection is probably a gross overestimate.
In order to produce the necessary steering force in any perpendicular direction, you need at least 4 engines mounted perpendicularly. Even if the average deflection over the course of the flight is infinitesimal, the engines need to produce enough thrust for the maximum thrust needed at any point in the flight. For a 12º maximum, the perpendicular thrusters would have to each produce about 20% of the thrust of the main engines! That's big -- for a Falcon 9, each of the lateral engines would be nearly as powerful as two of the main Merlins. For rapid response you'd need pressure-fed, hypergolic-fuel engines, which means separate and heavier propellant tankage; if you bring too little fuel for the laterals, you lose control mid-flight, and if you bring too much, you waste mass. With gimbaled engines, almost all your fuel goes toward acceleration; cosine loss at 5 degrees is less than half a percent.
Engines of that size don't have much in common with what's usually thought of as RCS. For rapid control you have to start and stop the engine quite quickly; doing pulsed hard starts like that on a 490N thruster is one thing, doing so on an engine 1500 times more powerful is going to be positively destructive -- and I'm not aware of any existing pressure-fed engines of that size.