it's best to learn on an engine that is not your daily.
compareing chips is easy done, then you can try each chip and see what goes best. thats about as safe as it gets.
when you start pissing with them checksums become a problem and you have to pay for the software that sorts them out for you.
but general AFR tuneing is easy and safe, shove a wideband lambda in your exhaust, datalog it vs manifold absolute pressure and RPM, find your fuel map and offset it up slightly say 5% and do the same run again.
this will give you a good idea how much the adjustment in the map relates to how much adjustment in the actual AFR. your looking for 12.5:1 at full throttel ( for basic mapping ) so you go back and adjust said map so drop it where you've gone past 12.5 :1 and increse it proportionaly where your still below it. then do another run and see how close you got.
if you sit there and work the numbers you can start out doing it a lot more presicely get a pita with eeproms tho as you have to open the ecu every time you want to tweek unless you buy an emulator and there expensive.
as for spark thats where the power hides but also the risk, on the fked old abf i used for testing i din't give a f**k so i just unplugged the knock sensors so it wouldn't back off the spark in it knocked and i just knocked the spark map forward in 2 degree steps till it knocked then back 3 degrees when it did.
it much better to do it on a dyno tho as there is no point adding spark if your not getting more torque from it. and you can see if drowning out the knock with fuel and adding more spark makes it go better or not.
