OK, some more clarification.
1. I don't have my old test data. I did those tests years ago. I also don't have any current cases with airflow issues but I can take some readings once I set up a system or two with top-mounted PSUs that don't have a top mounted fan/poor airflow. Not my top priority though. I originally did the tests on top mounted CPU cases without/with an exhaust fan or fans mounted in the top of the case. I did not switch to a bottom mounted PSU case for quite a while after that. It really did become a pretty moot point to me at least as good cases came out that actually had good airflow as opposed to the usual PSU fan and maybe an 80mm intake fan in the front of the case and if you were really lucky an 80mm or 120mm exhaust fan at the back of the case.
I'll do some testing as I am building systems and either post a separate thread here or on a web page with results. I am not planning on cutting up any cases to add more fans or exhaust holes, or relocated power supplies, so it will not be 1-1 results, but I can possibly leave the side off of the case or maybe even remove the top of the case depending on how it was made.
When there is air coming out of the PSU and the exhaust fan that is significantly warmer than the ambient air, you have less than ideal airflow through the case. I strive for, under full load, the air coming from the case exhaust to be barely noticeable in temperature by feel. I would have to do some actual tests for the exhaust air, but I would say maybe 3-5C difference from ambient is acceptable.
2. With a top mounted PSU and no exhaust fan in the top of the case or way for the hot air to escape, with the fan on the bottom of the PSU (almost ALL good PSUs are made this way now), unless you mount the PSU upside down, you are still going to have at least 3-4 inches in the top of the case where there is a pocket of hot air. And guess where the optical drives are normally mounted? Yep, in one of the upper 5.25 slots. Excess heat also kills those rather quickly.
3. Depending on the article about temps and caps, the range is from 7-10C for doubling capacitor life expectancy. It is also going to depend on the quality of the caps and the real temp rating on the cap. Crappy caps are all over the place, are often overrated for what they can actually handle, and have nowhere near the life expectancy that is given by the scam companies that make them.