Desk Top Dyno
The big problem with computer programs is that most of them use head flow data to predict power output. But Kaase, as well as others, have proven that head flow (as measured on current flow benches) doesn't equal power output. Kaase has some great discussions about how his Hemi heads make a lot more power than his wedge heads even though the wedge heads flow the same on the bench. A 400 cfm head can make 700 hp or it can make 900 hp. When that is true then the computer program can't make an estimate.
I witnessed this recently when we tested a GM 632 crate engine. The 632 has a hyd roller cam with beehive valve springs and it runs on pump gas. But the heads are so good that it makes over 1000 hp. The heads on the 632 are crazy, the ports are raised so high that you can see the intake valve thru the intake manifold. The GM 632 crate engine is basically a big pump gas shortblock with Pro Stock heads. The heads are so good that the engine can't help but make gobs of power.