Dave, great question. I was hoping someone would ask that.
With a project of this nature, I think the answer is always yes. Long story short, I now am a user of "Virtual Engines", a very powerful computer code for modeling IC engines. Not to be confused with "Desktop Dyno", Virtual Engines takes into account all the inlet and exhaust ducting, plenums, valve and port geometry, cam/valve events, combustion chamber surface/volume ratios, etc. The program will then produce power, torque, IMEP, PMEP, BMEP, delivery ratio, thermal efficiency (and more) plots that are very accurate to what a real engine would do on the dyno - sometimes within 3%. Armed with this powerfull tool, I've modeled my K301 engine as described in the article (a spare time project).
I've never actually dynoed my engine so I'm relying on the predictions, but I know from experience in modeling other L-head engines that I'm within 10% of actual test data. The calculated results indicate peak power of around 16 BHP @ 4200 rpm with the Madson cheater cam, and peak torque of 23 ft-lbf @ 2600, as built.
Here's what helps. The larger K321-341 exhaust valve is good for almost 1 hp increase at 4000 rpm. Adding a 22 inch straight exhaust is good for about 1.25 hp increase. Going to the larger #30 carb yields slight benefits above 4000, but detracts at peak torque. Advancing the cam timing by 6 crank degrees moves the power peak down to 3800 rpm where, with all the above mods, produces an amazing 18 BHP with peak torque bumped up to 24.5 ft-lbf @ 2400 rpm.
The straight pipe is not acceptable in the 'hood so next best is to gut the muffler and make a glass pack. Not as effective as the straight pipe but good for .5 hp increase vs 1.0 for straight pipe.
Now you know all my secrets and what I'll be building for KK#2.
(Message edited by dkirk on May 24, 2004)