You are absolutely correct that there could be an issue between the elevator servo and the stick, which would cause an aileron deflection to input elevator movement. The collector pulleys take out the interaction between the stick and elevator but not necessarily the stick to servo.
This additional sketch shows an aileron servo (probably looking at the wrong side of it here) that has its control arm output IN LINE with the stick's longitudinal axis. This would create a conical movement, without shortening or lengthening of the linkage.

Good luck fitting all three servos in their optimum locations in that little area however.
(Note that the linkage to the elevator servo is at 90 degrees to the servo arm - or you will get differential)