@kerberos_20
Where are you getting some of your info from?
Push, pull, push-pull matters very little. It has a greater impact on:
-thick boy custom rads, like 60mm thick.
-poorly designed front or top panels, where it's ideal to position the fans as far away from the panel as possible. With open mesh, it practically doesn't matter as much.
This here is something I believe top AIO users need to revisit: leave the rear slot empty, except for a filter. No exhaust fan, nor an intake fan.
I've tested cpu and gpu thermals with the hardware in my signature by running Cinebench R23 and Unigine Superposition(looped) at the same time, and between the 3:
-exhaust had the worst gpu thermals compared to the best result by 3C.
-no fan was the best.
-intake was within margin of error of no fan, but made a little more noise. I figure it doesn't do better cooling, because it's interfering with the pull from the AIO fans.
In my build, gpu AIO is top exhaust, cpu AIO is front mounted intake. I also DIY'ed a mesh window in place of the acrylic one.
Top mounted cpu AIO + open air gpu, which are more common, might be different, or similar. That's why I bring up the need to revisit.
You are not missing much between AIOs. Fans and ears = performance.
-Fans between models have different specs at 100%, but hardly anyone does that.
-Manufacturers don't post fan curves(save for 2), and performance below 100% is not linear; some models will do better/worse than others at a given range.
-Different ears, different perceptions of sound, different comfort zones, different distances between the user and their PC.
There are too many variables to tell you exactly how much one cooler trails/leads another. I think the biggest difference between the Capellix and Galahad is the mandatory software for the former.