There is no difference in the philosophy of trimming a swept wing or a straight wing aircraft; always trim in yaw then in roll. However, all other factors being equal, a swept wing has greater lateral stability (dihedral effect) than a straight wing. Therefore, any sideslip that is present due to directional mistrimming will generate a greater rolling moment with a swept wing than with a straight wing, and this will appear as a greater out of trim condition in roll.
One aspect to consider is the definition of 'in trim' in yaw. Typically, the slipball (or other lateral acceleration symbol) is trimmed to neutral because that is the primary indication. However, note that in wings level flight this may not be zero sideslip, especially at high power settings in a single prop aircraft or in a twin prop when both propellers rotate in the same direction. This is due to a lateral aerodynamic force due to an asymmetry in the airflow over each side of the fuselage and fin which, for equilibrium, must be countered by an opposing aerodynamic force that is generated by sideslip. Similarly, if you are in wings level flight and OEI with the 'slip ball' centred you most certainly will not be at zero sideslip and the resolution of the forces here follows the same logic.
There are some aircraft and some circumstances when you may wish to maintain zero yaw rate/constant heading but to do so the slip ball will not be centred. Typically, a decel to the stall in a prop aircraft or the case of an aircraft such as the Buccaneer which had the speedbrakes extending into the jet efflux such that unless they were perfectly symmetric a yawing moment would be generated but with little effective lateral force. The Bucc was flown in the landing pattern and on the approach with varying amounts of speedbrake and you just trimmed to constant heading and ignored the slip ball.