Here's my contribution:
http://smharr4.dnsalias.net/cars/93300E2.8/001183.wav
Engine was cold when this was recorded with air temperatures at 37F. Recording device was a iPaq 3800 placed on the battery, mic pointed towards the engine. Car has 166867 miles on the clock. ACC was off, as were the lights (except the interior dome light)
The engine needs motor mounts, and I think the air pump is failing (you can hear it drumming about half-way through before finally settling down). Newly replaced items in the last year include blower motor, timing chain, tensioner, rails & guides, fan bearing, idler pulley, fan blade and viscous clutch (2005 was not a cheap year for this vehicle).
I think this could be useful, but the variety of the quality of recording devices could make this tricky. It would be great if we had consistent recording equipment, and could identify individual failing engine components just from the unique waveform they make on the recording (a typical high-tech solution from someone whose employer sells supercomputing time for applications such as these).