PSU output ripple tests under a range of loads (& crossloads) are a useful test for design stability. An example has been the FSP Epsilon series, where the output ripple characteristic has shown the problem of its design. In most cases the ripple signature is within the spec, but doesn't look great. Care needs to be taken when assessing "ripple". It's fairly common to employ ultra low ESR output caps on the secondary side which can initially give nice results. The problem is that most use aqueous solution electrolyte that after perhaps 6 months starts to run into problems with stability in the PSU environment. I'd certainly prefer higher load life non-aq caps with their inherent higher output ripple as long as the signature is that of white-noise and does not exhibit a pulse signature...