One important thing you must not forget...
A lot of the "FX" like audiosnap and V-Vocal as examples have at least 2 or 3 quality settings.
When you hit the Play button to hear your mix in real time Sonar will default to using lower quality/"faster computing" settings in order to keep up with all the calculations required.
When you "export to audio" the high quality settings are used by default as there is no "real time" requirement, so small anomalies you may hear from, say, v-vocal will dissapear in the exported version.
This does mean that in practice the exported version will sound slightly different to what you hear when you simply press Play, but hopefully in a better way.
There is a way around
some of this, and that is to render individual tracks first to force using the high quality settings, (bounce to track), though you still have the same issues with FX on the Master bus. Rendering the Master bus is, by definition, exporting to audio...
It should be noted though that none of this is affected by the quality of the soundcard as the soundcard simply converts the final calculated data stream to audio.