I think Spoolin and Nuke are both right, alone a P&P won't net much. But when part of a thorough engine development plan itll be the icing on the cake. OVB has a good write up you might want to check out -
http://www.overboost.com/story.asp?id=431
With so few performance engine parts available, and so little development done for the Gen 3, somebody has got to cut into it sooner or later. And in the end the fastest car, everything else being equal, will be the one with the best headwork. Period.
Right now I think headwork is going to be the easiest and cheapest way to raise compression and make good N/A power. If you bugger up a head its no biggy either, because obviously these cars are already showing in the junkyards. The only things missing are cams, ECU tuning and a header. Crower can do the cams for about 5 bills. The ECU is just going to be a pain, but it is reprogrammable. A f*ing header is the only sticking point for now.
If youre going for it, look into extrude honing too, its kind of like a P&P is one easy step. Its also the best cheat going in SCCA show room stock class cause it needs no other mods to back it up. But if you also deck the head slightly to raise CR, and multi-angle the valve seats youre done for about $400.
The important point is to have a plan and know exactly what youre doing. I think good results can be had for about $2K with proper knowledge and planning.