yea im gonna try something like that. But the problem is that the screws transmit sound to the case too, so you need to wrap the screw with something so there is no hard contact. Might be worth it to just buy grommets for everything in the long run.

Wal-Mart has 8.5x11 sheets of 2mm thick foam dirt cheap in the crafts department, both glue-backed and plain in like 10 colors. Easy to cut to shape, fold/layer. A little too compressible. The only thing I worry about is deterioration over time.

Also bought 3 of the Directron 120mm octagonal fan frame grommets/isolators with my last Panaflo 120mm order. The molded "silicone" (silicone rubber?) is pretty hard to compress, not too much give at all, but probably will do OK. The trouble is that it extends a little (2-3mm) into the airflow path. The "silicone" may be cuttable without fraying, I'll try soon.

Radio Shack sell pretty cheap baggies of assorted screw grommets, by the way

The only problem with grommet mounting a PSU is that because of the weight, the screws may just pull right through the grommets. Unless of course your case has a ledge or bar to support the weight of the PSU (suitably damped of course!).

For mounting heavy items like PSU, or fans hunt around the studio
acoustic soundproofing suppliers for offcuts or pieces of sound barrier mat.

It's a rubber sheet about 2-3mm thick, 5kg/m^2 (very heavy).

o Mass - so eliminates resonance & low frequencies
o Isolation - by compliant-enough rubber base
o Gasket - durometer is low enough to get a good seal

It's cheap stuff, about 15$/15ukp for 2-square-metres,
but much more effective as a gasket than the feeble foam.

Dynomat is very poor in terms of gasket capability and is
frankly too overpriced for this application.

If really desperate:
o Cork material cut into a gasket - long used in vibration isolation
o Vinyoleum, altho that is getting desperate - works tho

A poor mans resonance dampening for cases/enclosures is normal
"Treadair" ribbed rubber closed-cell foam-underlay. Not that cheap,
but very effective at dampening noise & vibration of even big servers.
The problem is finding a contact adhesive strong enough - most are not.

Rubber grommets will not work satisfactory, and may introduce a short
circuit path for air reducing the effectiveness of case cooling considerably.

Foam alone will only absorb high frequencies, mass is needed on thin
sheet panels to prevent resonance. Treatig causation is still the best route.

To add, the PAPST neoprene foam gaskets are very poor, the rubber
moulded gaskets by comparison are extremely effective but big $$$.