Drive it for six months and note what you dont like about the car.
Figure out how to quantify that particular feature
Decide what the new improved value of that feature should be when you are done
Perform modification
Measure against baseline
Cry when you realize you made it worse. ( as happens with a lot of mods)
OR
Start with no measurable baseline and go with "Mad Tite JDM Mods cause Racecar Brah!"
Both methods seem to work
