Lol-ed from his skills in tweaking those obscure items, i'm sure we will have a LYN-optimized setting soon...
on the sampling front, i wonder if taking more samples means (directly linearly) affecting battery life (similar to a person speaking to many other people in a survey?) or does changing the sample size from the normal 30 nos to let's say 10, result in a direct (although microscopic) battery savings of around 66% (for that particular task)??
or is simply UV-ing it a more "wholesale" approach to battery savings
there's only so much you can UV until your chip (which i believe the tolerence / quality of each and every chip is different - read from AndreiLux @ xda IIRC) cannot be stable at such low voltages... so i think these small but significant changes (besides the criteria for core on / off) will help in the overall reduction of battery usage!
then we will have a highly efficient (not a battery savings profile but not a blood sucking performance based one) settings profile that will be easy to use and follow?
i wonder if stweaks has a profile that you can save and distribute so that others may tweak it more easily (instead of manually pushing those "knobs" on the cpu and gpu especially).
another thing, do you think that messing about and tweaking the memory has much help in performance / efficiency?
/data/.perseus/default.profile --> this is the file for your settings in stweaks with perseus that other can use directly by placing in the same folder and reboot.
best if you remove any UV if you wanna distribute that file in case the UV does not work on other device and ended up cannot boot.