Interestingly, since single-plane 4V intake manifolds are in common use, we should re-mention that they are inherently provided with two basic (and of different length) runner designs. While much has been said and explored regarding how to optimize these two sets of runners (camshafts ground for boosting torque at two different engine speeds, exhaust systems designed to complement both runner lengths, different rocker ratios, and so on), the fact remains that exploiting such methods is a viable way to "integrate" these types of manifolds with companion parts.
A similar approach can be taken with respect to matching exhaust systems to a particular engine displacement and intended rpm range. Although we are dealing with a "dry flow" system (as compared to an intake system), differences in working fluid temperatures (air/fuel charges vs. exhaust gas) and piston position when peak flow rates are generated (roughly b.d.c. exhaust cycle and mid-stroke intake cycle), we can apply the same approach used for intake manifolds. If you missed it during earlier discussions, the calculation equation is as follows:
Peak torque rpm = (pipe or passage cross-section area x 88,200) / cylinder volume.
You can certainly perform some algebraic operations on this equation to solve for a required pipe or passage cross section area or cylinder volume (one cylinder).
Again, as with intake manifold runner length, header pipe length changes (longer or shorter) tend to rotate a given torque curve about the peak torque rpm point. Shorter pipes increase torque above the peak while decreasing it below this point, longer pipes increase lower rpm torque and decrease it above the peak, all else being equal.
Before we leave this little math model, there's one more instance where it can be applied. Let's say you have some engine dyno data for a particular engine that includes a full torque vs. rpm data stream and that you've measured the cross-sectional areas of the intake manifold and exhaust header passages (primary pipe only). By comparing the dyno data's peak torque rpm points with what you've computed and averaged for the intake and header system used, you can determine if the engine is "over-ported" or "under-ported." For example, if the actual peak torque rpm point is higher than what you computed, chances are it's an over-ported combination of parts. And, of course, the opposite is true as well.
Moving on to camshafts and armed with a specific range of targeted engine speed, you'll find that most camshaft manufacturers have a wealth of experience and technical information to help in the selection process. Just be realistic in your expectations about the rpm range for which you plan the most on-track use.
Cylinder heads? Not so much because intake and exhaust port path lengths are comparatively short (as measured against intake runner and exhaust pipe lengths), it's the need to help make transitional and directional flow changes into and out of the combustion space that cylinder head ports are believed to be important. There are indications that providing efficient and effective transitional flow into and out of the cylinders is more important than port tuning.
Some thoughts about practical applications
As you're probably aware, there are several very good sources available today for PC modeling of engine component packages. Their use will enable you to circumvent many of the problems often associated with sorting out specific engine parts combinations. Considering the costs associated with "live" testing (either on a dyno or at a track), these packages are certainly a cost-effective alternative. The benefits are even more pronounced when you consider the expense associated with full-on CFD and other such computer models.
At least from experience, regardless of the analysis method you choose to use, it's critically important to make your best estimation about what engine speed range you want to target, well in advance of any steps to determine the best combination of parts. The results often trail down to racing cost reductions and more black and white flags hanging in your shop.
For the most part, a successful circle track engine relies on torque, whether it's for improved corner exit speeds or helping acceleration past the flagstand