I've just seen my fair share of it. I used to work in a type testing lab that did quite a bit of EMI testing.
I've seen what it takes to shield cables carrying digital information from EMI interference. Test results could differ from one run to the next without changing a thing at all. Shielding is usually done with a grounded plane (the tinfoil), but it has to be connected to the power ground in multiple spots with very short wires. You're trying to provide a path to ground for the interference before it finds its way into the digital signal circuit.
Besides, the brain isn't the only thing that needs shielding. All wires entering and leaving the ebox must be shielded.
Because of this constant variance in shielding results, the best solutions are cutting down on the source of the EMI and increasing distance between the source and the circuit.
In other words: You won't be able to come up with decent, waterproof, consistent shielding for a watercraft ignition. Reduce the noise at the source and try to increase distance.