I have gone to great lengths to self-master my various CDs out there in the void.  I have had AAMS for a number of years but never used it because I didn't understand how to use it ... because I never read the manual.  Duh. 
 
So, anyway, being bored, I read the manual couple times and watched the videos.  Listening to what it does against the various "references" it seemed that the "modern.aam" reference worked best for my stuff.  Choosing to use the 50 band EQ and 8 band compression, I fed it a few of my already "mastered" mixes and was very pleased.  Once I looked at what it had done to the EQ, I saw that I did have a bit too much low end in some of my tunes, etc.  Bottom line is that I agreed with what AAMS was showing me as deficiencies in my technique, and that the result sounded better than it did before processing.  So I'm kinda blown away I've had this and never gave it a try.
 
For example on my self-mastered song "Always Me", all of the EQ bands consisted of tweaks in the 1 - 2 db +/- range which seemed to indicate that I had generally done a fairly good job, and that AAMS was not going EQ-batsh*t.  However, it did choose a 3.7 boost at 56 Hz and a 4 db cut at 103.  So I listened closely to the 2 versions, and I had to admit that the AAMS version had a cleaner sounding low end, and that the kick stood out tighter and more clear.  The compression across the 8 bands was a gentle 2:1 with 1 to 3 db compress which is what you might expect from mastering.
 
I helped construct a million dollar studio (that immediately went bankrupt), and have been recording more years than I would care to admit.  I've treated my room to handle early reflections, and have bass traps in the rear corners.  I have sophisticated and costly plug-ins.  I've studied to make myself better at what I do.  Both me and my wife (who has perfectly flat 20 to 20,000 hearing) think that AAMS re-masters of my masters sound better than my masters.  Damn.

Greg Raves