While making these changes can result in a major difference in Xbench, don't get too hung up on them......my tests with real world apps show a system with beam synch disabled is not reallly faster. In fact, it makes me question the validity of Xbench tests, at least on PC. Too much weight is given to screen redraws maybe? Also, beam synch helps to smooth video windows, etc, and prevent "tearing" when moving windows. Obviously, beam synch disable can't improve processor speed, etc. As for the Quartz 2D, I've tried it with mt Intel GMA900 graphics and it's quite buggy.
I've been experimenting with my setup a little, and it seems that disabling beam sync has no real-world effects on anything - the only difference is "tearing" and some bugs in DVD playback. It seems that there is no other reason to disable beam sync than getting impressive XBench results...
I've been unable to find any bugs with Quartz 2d, it seems to work just fine and it does make the GUI a little faster. I've got GMA 950 graphics and 10.4.3 system.