From my own observations of unrest/income relationships, I believe the way they are related is logarithmic in nature, in an inverse relationship. That is, as unrest increases, income decreases, but the reduction differential always decreases (towards 0). I've noted that, after taking a rich province with high unrest, I still get a good proportion of the province's 'normal' income (income with no unrest). OTOH, even low amounts of unrest can reduce income significantly. These observations have led to the above hypothesis.
What the modding parameters would do, under this system, is move the point where income is halved, but the effects on either side would not be as affected.
It'd be much easier to describe this graphically, but I don't know exactly if I could do it here.
Anyway, I hope this helps.
