Taking all the information here in forums into consideration I was able to look at my case in a different light. I figured that if I ducted the HSF into the PSU intake, which is very close to it, I could vent the major heat producer quicker and by boxing in the PSU/CPU area bounded by the video card (as I've seen here in the forums) I could get the heat out of the box quicker. I could keep the case heat from affecting the PSU and CPU, keeping the PSU fan speed down. My case has a spot for a front case fan with some vents in the bezel and in the rear above the cards is a long vent. I thought that if I isolate the "hotbox" and keep the case fan blowing inward, that the flow would go over the cards and out the rear vent. It has an almost direct path. I also think that this arrangement cuts down on turbulence and competition between the HSF and PSU intake, which are 20mm apart.
I made a made a duct enclosure. Paper encloses the area between the HSF and PSU intake cutting the case off from the PSU. The case is vented only by the flow from the front case fan exiting the rear vents above the cards. The MB temp has been lower than usual at 31C, usually 36C. So the PSU does not get the heat load from the case.
All the heat load of the HSF is vented through the PSU, which is the most direct path.
Immediately the sound of the HSF was reduced by about 90% with the flutter and chuffing inaudible outside the case, just a lower whirr. The high frequency sound from the HSF was gone. The PSU now was predominant. I was surprised that an enclosure made only of lightweight laser paper could do.
The I-Storm was not powered, but still in the bay. This does affect temps of the HD slightly, lowering from 34C to 31C if I take the hood off. Closing this off completely would raise the HD temp.
I monitored the flow out of the PSU with a remote thermometer. It stayed at 31C steady.
HD temps were much lower until the case heated up a bit after 30mins, about 36C, but was in the high 20s for a long time.
MB was very steady at 31C.
CPU was lower than usual then rose to 47-48 under light browsing load.
PSU fan was much lower than usual for much longer, rising slowly to 2200, which I think is less than the usual top rpm. It was going back and forth between 1900 and 2300. It was 1700 for a long time, very unusual without the duct.
The PSU exhaust feels warmer than normal, but I did not take a reading before adding the duct.
up to 32.9C after 30min, 1hr
HD holding at 36C after 1hr
Ambient room temp measured on top of the case front 27C (winter).
the MB temp is lower with the duct from 47-49 depending on the ambient temp. Now it is topping out at 48C when it normally would top out at 47-48C all day.
although the CPU fan is not temp controlled, it does vary with heat, from 5800 to 6800 normally, but with the duct the speed is staying at 5800
Conclusion is that more fans is not the answer, ducting reduced fan speed and temps from the highest, although probably did raise the CPU temp a bit. Also, the vent holes down the side of this case likely contribute nothing but a sound path and unhelpful flows. It does show that dividing the case into differing heat loads helps.
I'm certain this case does not have sufficient airflow because when the cover is off temps dropped by one.
Some pictures.
http://www.city-gallery.com/knoblock/a500duct-1.jpg
http://www.city-gallery.com/knoblock/a500duct-2.jpg
Sorry if this is rambling, I've just done the experiment, Opera does not seem to like BBCode buttons, can't seem to link the pics, have a cold.