I've seen a few so called "Passively cooled" processors around here and have been somewhat perplexed by their owners' proclaimed passive cooling system. It seems that if there is no fan directly mounted to a heatsink, then it can be considered passive - even if it is a mere 2 to 3cm away from a case fan!
Similarly I have seen "silent" systems which clearly have not only a mechanical hard drive, but a number of fans too! I propose the following formalization of silence categories:
No forced airflow over the component via directly attached fans, and no or negligable airflow induced from nearby fans. I.e. a vague draft from an ATX-spec cooling design is OK, but with the heatsink next to a high airflow area such as an exhaust fan is not.
No fans (or water pumps) whatsoever. Not even a case or PSU fan.
A system containing few or no fans or pumps, but with one or more moving parts (mechanical hard disk, generally accepted quiet fans, etc)
As above, but with no moving parts whatsoever (No mechanical storage), also, edge-case noises such as coil whine must be mitigated significantly. Further, optical drives (if any) should be a generally accepted quiet model. Finally, if / when ion coolers become popular, any turbulent airflow noise from such forced convection systems must also be mitigated.
In any case, the system must of course be stable and reliable. If any components have been modified such that their performance is reduced (E.g. underclocking), this should be made obvious to readers.
From the above there could be a few different system descriptors:
A Quiet System with one or more Passive Components
A Quiet, Passive System
A Silent Passive System
I hope my step in the direction of standardization does not come off as elitist (cue comments about my post count), rather I hope it sparks some interest in this area in terms of more accurate objectives in silencing as well as descriptions in communication.