Famous quote from A.J. Foyt: "Speed costs money. How fast do you want to go?"
It's no secret that building a Pontiac is very often more expensive than building a bowtie -- owing to parts availability, parts cost, and difficulty in finding people who actually know what they're doing. Balancing the money side of things is the hardest part and there are tradeoffs always. For example: building a flat tappet engine is a crap shoot these days because it's now literally been decades since any "new" cars had flat tappet valve trains, so the quality of parts, especially the "affordable" ones has declined. You can do everything right including using high zinc content oil (if you can find it) and still just be unlucky and wipe out a lobe, causing you to have to start over. The obvious answer is to go with a roller system, but those are more expensive in terms of parts cost and also indirect cost (different valve springs, different installed heights that require machining the spring seats in the heads, different pushrod lengths, etc.) -- and they too come with their own dangers. If a roller bearing or bushing goes a way or a link bar fails and allows the lifter to turn sideways... the results ain't pretty.
Building an "all out" Pontiac and doing it right with the best quality parts can easily cost you north of $10k. I"m not trying to discourage you from doing that, in fact quite the opposite --- however it's sad when someone starts down that road without realizing what they're getting into and then have to stop without ever getting it finished because it got too expensive. The result is another permanently dead Pontiac - and no one here wants that.
So, back to your question: First be honest with yourself about just how far you want to take it and realistically how much you're able to fund the project to get it there. Details like gear ratio, transmission type, rear axle type, how you plan to use the car (street machine, race car, highway cruiser, etc.) all influence your other choices. An all out race engine that doesn't even begin to make significant power until it's north of 4000 rpm is going to be a miserable experience in a car that spends most of the time cruising the streets down to the local hangout, and it's also not going to be very successful in "stop light wars" - or anywhere else except the drap strip.
Ok --- heads. Just about any of the popular aluminum aftermarket heads are going to outperform all of the Pontiac factory heads, except MAYBE the rare as unicorns Ram Air V's. I personally have a set of ported Ram Air IV #722's that I had flow tested, and even those don't flow as well as my Edebrocks that have been treated to a basic entry level street porting. The tip on aftermarket heads is to buy them bare - i.e. without valves or springs. There's a reason that the price difference between bare heads and so called "ready to run" heads isn't that much. Check out the videos:
No, buy the heads bare and have them sent to an expert "head guy" to have them outfitted with high quality valves, springs (to match the cam you've selected), retainers, locks, etc. If you can afford it, spend a little extra to have him work the ports a little.
Camshaft: The most significant criterion for selection is expected RPM range. Where is the engine going to live most of the time? Pontiacs aren't chevys so we don't have to fret so much about killing bottom end torque, but still you can't ignore that completely. An engine that has a torque peak well within the RPM range where it spends most of its time is going to be the most fun to drive. Don't go chasing horsepower to the exclusion of all else. Remember HP is really just a mathematical formula: HP = (torque X RPM) / 5252. (That's why on every honest dyno sheet the torque and HP plots will *always* cross at exactly 5252 RPM). Take an engine that makes say 450 lb ft at 2500 rpm vs an engine that makes the same 450 lb ft but at 5000 rpm ---- the second engine is making TWICE as much horsepower (because of the formula) even thought it's not pulling one ounce harder. Furthermore, on the street and everywhere else except maybe the drag strip, that first engine is going to eat it alive (because at normal street rpm, engine #2 is going to be a dog). Pick a cam that's going to put the torque peak in a reasonable spot, then work "outward" from there.
Compression: You can get away with more with aluminum because the material doesn't retain heat as well as iron, and heat is one of the main contributors to detonation. In fact, because of this you really NEED to up compression when you swap to aluminum. This is an area where you'll find passionate opinions all over the map about how much is acceptable, "dynamic" vs. "static" compression, ad-inifinitum. I have my opinions too. However consider this: On a very healthy Pontiac 461 with good aluminum heads and a moderately aggressive (for the street) solid roller valve train, raising the compression from 10.5:1 a full "number" to 11.5:1 makes very little difference. How little? About +15 in both torque and HP -- in an engine that was already making north of 560 in both torque and HP. Taking it from a safe on 93 octane (for aluminum) 10.5:1 to a risky for detonation 11.5:1 was good for about a 2% increase, and that's ASSUMING that you can keep it out of detonation at 11.5:1. Worth the risk? Your choice, but personally, not for me, not on the street... not unless I'm trying to milk every last bit out of an engine for race purposes - but in that case I'd be building it to run on race gas anyway. For iron heads, knock off "a number" - on 93 octane pump gas personally I wouldn't push above 9.3:1-9.5:1 tops.
My .02 - your mileage may vary - void where prohibited by law - slippery when wet - past performance is not an indication of future return.