Regarding these new tests from Phoronix, XanMod kernel overperforming Liquorix kernel in different workloads (except gaming tests, where Liquorix is around 5% ahead).
Afaik, Zen kernel based on Liquorix with additional tweaks, but I can't find any relevant comparisons of these two.
Has anybody already performed some fresh benchmarks and can comment on this topic?
We used Linux-tkg kernels in the past and had a lot of cases where hardware would not work as expected. This is probably caused by custom kernel configs (they are different than zen/default Arch). Therefore, to ensure compatibility and working hardware, we settled down on Linux-zen as it is an officially supported kernel on Arch
Personally, I didn't compare Xanmod and Zen kernel configs, but I would strongly assume it's similar to TKG kernels (highly tuned and deviating from the official one).
XanMod is not mentioned in that thread, but I did notice there are five different versions of the XanMod kernel in the Chaotic repo. That makes installing the kernel just as simple as installing any other package (no need to compile your own kernel for testing). You can throw a kernel on there, check it out, and just uninstall if you decide you aren't going to use it.
If it were me I guess I would start with these:
chaotic-aur/linux-xanmod-edge-headers 5.18.11-1 [29.51MiB 126.68MiB]
Headers and scripts for building modules for the Linux Xanmod - Latest Mainline (EDGE) kernel
chaotic-aur/linux-xanmod-edge 5.18.11-1 [90.44MiB 348.86MiB]
The Linux kernel and modules with Xanmod patches
Why not install a few kernels and perform some of your own benchmarking? The results you get on your own hardware will obviously be more personally relevant to your usecase than what Michael Larabel gets on his latest test machine, or for that matter anyone else using a hardware stack that is different than yours.