The engineer using Pro Tools was obviously working with a lot more than just normalizing.
And yes a louder track will always sound "better"
Some of the comments here are from people who may have not read this thread carefully,, so I'll say it again on behalf of those who have already explained this very clearly...
Normalizing is not at all like compression. It is almost identical to the "gain" function with the difference being with gain, you select an amount and apply it. You can add or subtract db level.
With normalizing you set a limit like -1 db and apply. You automaticly know what will happen.
With gain you risk going over the top unless you know what the original peak level was. You certainly don't want to apply +12db of gain to a track that is already -8db.
Most wave editors have a peak level analyser so you can make an intelligent decision on whether to normalize.
I have done years of working with stereo files of everything from live recordings to tape and vinyl transfers.
All it takes is one spike in the track and normalizing will do you no good. A low level recording with spikes through out will need manual removal of the peaks first. Often a compressor is just not fast enough. You really need to understand how it works and when to use all the tools that are available.
And this brings up the topic of needing to understand and use Average RMS levels( or the other new one)
Even with manual editing of peaks and normalizing your track might end up at only -20 db Average RMS.
Cakewalk mumbled something about improving Sonar's wave editing,, I for one would love this more than any possible other upgrade. I live for wave editing, and once you used a good one,Sonar's options are way under par.