I'll do us all a favor and keep an open mind about this.
Perhaps the first question we should ask "generally" is "Which temperatures are we measuring, and how are we measuring it (them)?"
It has been my experience with a heatpipe-tower cooling strategy that a fan deployment of ample intake CFMs -- concentrated so as to exhaust only or mostly from the cooler and rear exhaust fan -- will keep CPU temperatures lower. But this also assumes the cooler exhaust is ducted to the rear fan, not allowed to mix with case interior air.
If you position thermal sensors around the case, it may be that the closed case has higher temperatures -- I don't know for sure.
One experiment to improve cooling from a single-fan AiO cooler here in the forums, using a small midtower case like a Corsair C70, proclaimed lower CPU temperatures for a closed case than for an open case. Since the case was smaller than my own midtowers, the opportunity for 140mm fans had to be thoroughly exploited, and I think they were deployed as intake wherever possible, leaving a radiator ducted on either side with pusher and (exhaust) puller as the sole exhaust port in the box, except for some leakage through the PSU or some drive-bay vent-cover that wasn't blocked off.
I think the difference in the C70-type case for open side-panel versus closed was measured at something between 2 or 3C.
I personally block off all vents that only serve a passive purpose -- the ones without fans. I try to block all the holes, but I've left the bay-cover vents unblocked to assure some airflow across my optical drive or hot-swap bay. The more selective I am in choosing these unblocked vents, the better the pressurization -- either more air molecules or faster air molecules would traverse the heatpipe fin array. And the more effective the cooling on the bay components under consideration.
It would seem that the smaller the case, one could use either fewer or smaller fans, but choose fewer and larger where possible -- specifically for intake. I use an exhaust fan that may still be only capable of half the rated CFM for the intake fans, but then -- it sits behind the cooler and isolated exclusively to the cooler because of the duct-box. But even for the smaller case size of a C70 (versus my HAF 922), the airflow over the components would be more effective: Narrower spaces; higher velocity; greater pressure.
These are just some loose thoughts about this. So for an answer, closed-case may offer greater chance of lower CPU temperatures; open-case may show lower temperatures for a thermistor taped to -- say -- the motherboard, a drive cage, etc.
And I can only explain the former firsthand. Why would I run my computer with an open sidepanel, except to inspect it, examine the motherboard LEDs, check that fans are running, etc.? If you did everything possible to isolate noise and vibration, the closed case does better at muffling air-turbulence.
I've seen a lot of different case designs, and there was once a case I saw made almost entirely of perf-aluminum and aluminum frame parts. It was a metal case, but the entire assembly was one big fan vent. I don't recall the customer-reaction to dBA levels. I just haven't seen anyone sporting that case for years.
A friend has a lian li made, Rockfish branded tower case, and they have the side panel open all the time.
They claim that having the panel open drops temps better than having it closed.
His case has 1 120mm front fan, and 1 120mm rear fan.
For more pics check out
http://www.dslreports.com/forum/r20090312-New-Case-Rocketfish-RF-FULLTWR-Lian-Li-special
I know on my case, internal temps go up if I leave case open, since airflow is not going over everything like it normally would.
Case open, his HD temps are 39C, and closed, they are 43C on a typical 83F day with no AC.
I am wondering if anyone else has seen temps lower with the side panel open vs having it closed?
Facebook
Twitter