Well, I've seen worst in terms of ripple.
Voltage fluctuations though are pretty big. Revving up a CD-ROM drive it goes as low as 11.65V and once the drive is up to speed the power supply overshoots to as high as 12.32V. Definitely not within ATX specifications.
Edit: oops, those spikes in the scope capture are bigger than I thought. I forgot to set the scope to X10 since I'm using a X10 probe. So those spikes are actually on average 487mV not 48.7mV