Firstly, it would have been better to start a thread of your own instead of waking up a five years old thread.
In the struggle to get a "commercial" level final wav it does not matter what you hear, because what you hear depends on the position of the volume knob. You have to look at the level readings. What you burn on a CD surely is exactly as loud as the original wav, there's no volume adjusting involved in the burning process. Also, comparing the volume of a CD played by Windows Media Player to that of the project played back in SONAR is meaningless.
The perceived volume is most accurately shown by the RMS volume level, which is the average. IIRC the RMS of around - 6 - -12 dB dB is somewhat OK (others please correct if I remember wrong). SONARs track meters can show RMS as well, IIRC. The peak volumes are practically meaningless. High RMS levels are reached for example by
1) Volume automation on individual tracks when necessary
2) Compression and EQ on individual tracks
3) Compression/EQ/Limiting according to need when finishing the mix and exporting/mastering