This is why I like ths idea of an extra top exhaust, even if it is usually seen as a disruptor to airflow.
With 4x intake fans I do want enough air being exhausted and this is 10x as important with a GPU like a 3080, which pumps out an insane 350W of heat that needs to get out ASAP.
I think I'll be keeping the 2x top and 1x rear exhaust setup, run these 3 exhausts at a silent 45-50% rpm and run the 4 intake fans a bit faster at 60% or so. This seems to give excellent airflow, excellent results, and very low noise levels.
It is physically impossible to flow more air in than out / out than in. While stacking fans as intake on one side and exhaust on other generally doesn't make much if any difference, it sometimes changes flow channels / paths thus supplying cooler air where needed. But as pp111 says, it's 'a bit of black art'. It takes time to monitor airflow changes by recording air temp in different places while changing component load and/or fan / fans speed, then comparing data to determine what will hopefully improve overall performance with lower temps and noise levels.
Airflow is a bit of a black art. What works for one case can fail miserably in another. So many factors play in the final outcome.
In general driving air in to the case is great, but then it has to get out. The only reason you need an exhaust fan is not to exhaust the air as such, because it will happily find it's own way out, but to ensure that you set up a flow over vital pieces of equipment. No good if all your air goes out the bottom of the case and only a fraction is traveling past the CPU for example. The problem with front fans mixed with top fans is that you might be drawing air out and creating a flow that is actually bypassing everything important. Especially the top fan at the front.
Exactly right!
For me the best way to setup case airflow is monitoring air temp entering components I want running cooler. They playing with fan speed and placement until I get best temps and noise results.
What I've found is almost all tower builds work best with front full on good pressure rated intakes, all opening in front (maybe front half of case) not filled with fans block off so air fans push into case has to flow on through case, not leak into front and go in circles. Almost always removing all PCIe back slot covers improves back vent area, thus improve front to back airflow, thus lower fan speed and lower temps. Sometimes a bottom intake will help supply cool air to GPU, but need to be careful it's not pushing too much air in with can end up pushing GPU heated exhaust up into path of CPU cooler's cool intake airflow.
Best low cost remote sensor thermometers I've found are the cheap indoor/outdoor digital ones with wired remote probes/sensors. If probes are metal I tape them up so they can't short out in running systems. A cloths pin on a piece of stiff insulated wire can be twirsted / taped to probe lead so probe can be positioned in middle of airflow into cooler fans.
Then monitor and record temps and fans speeds adjusting fan (both case & component fans) speeds until optimum balance for best airflow with low noise and temps is found. Usually I run out of patients settling for what is expectable rather than optimum. Doesn't really matter if it's optimal if it's acceptable.
