In general, PCR is useful anytime you want to isolate, amplify, or manipulate relatively short stretches of DNA.
Do you want to isolate a gene and knock-in a mutation that is known to cause a genetic defect so you can study it? Do you want to tag that gene with a fluorescent tail? Do you need to stitch together several shorter DNA components into a single unit (perhaps joining a zinc finger to a nuclease and creating your very own targeted DNA-cutting enzyme)?
If so, PCR might be for you.
Tito
The third might be a bit harder to make a commercial kit for while steering clear of patents. It's also not clear to me what kinds of samples you'd be able to test. An ear of fresh corn would probably work, but what about (un-)popped popcorn, or cornflakes, or corn starch? It would definitely take some experimentation.
To make these kits really user friendly, ideally you could figure out some quick, easy way to determine the result of the PCR without running a gel, which require a relatively high voltage DC power supply and usually use Ethidium Bromide (a chemical that, depending on who you believe, is either nasty or really, really nasty) and UV light. I'm thinking maybe some kind of spectrophotometer-like setup, which are used in things like a NanoDrop, but I don't know how easy/economical it would be to build a DIY-level work-alike.
Pretty sweet!
Tito