Currently, with thusters having lateral thrust thrusters have their maximum thrust in a direction that isn't the primary direction, which is counter intuitive.
This is most significant with the current small thrusters due to their high lateral thrust. Reducing the lateral thrust will make the issue less significant, but it will remain the case that thursters will have slightly higher thrust when a ship flies slightly angled.
Realistically, thrust vectoring would allow aiming thrust through some range of angles. Importantly, thrust vectoring increases lateral thrust at the cost of reduced thrust in the original direction. The total thrust output remains the same only the direction in which it is applied is changed.
So instead of having lateral thrust I suggest that a thruster should have a range of angles over which it can apply full thrust. For maximum flexibility it should also be possible to make a composite thruster where a part can have multiple thrust entries, each with its own thrust and angle range, for the purpose of creating a part with multiple thrust outlets.
