[Christ]

james - the resistor thing would cause it to lean out, but it wouldn't add "filler" to the combustion process... that's his point - if you add CO2, it displaces X volume that now cannot be filled by air/fuel.

His idea is bascially water injection with CO2 instead.

Here's the problem - aside from cost, which has fervently been pointed out already, the rate at which the CO2 sublimates would increase with ambient temperature... it would nearly be impossible (given your theoretical setup) to keep track of how much CO2 you're adding to the combustion process, and you may even add enough of it that there wouldn't be enough room left for the air and fuel to create a strong enough combustion process to actually keep the engine running. Chances are, after the engine warmed up, you'd starve it, and it would shut off.

Another issue with doing this - you're making the air/fuel mix very sparse in such a large area, which makes it harder to ignite... ideally, you want a richer mixture nearer the spark plug, and a leaner mixture everywhere else, or at least a homogenous mixture everywhere.

Adding random (or even controlled) amounts of CO2 to the mix, if you add enough, will make it so that parts of the a/f mixture are all over the place. You'd have to control the timing for the gaseous CO2 to be injected, so that it would limit access to the areas closer to the piston, and air/fuel could be drawn in on top of it... what you'd have essentially done is lowered the effective stroke of your engine, more accurately, lowered the displacement... there are easier, cheaper, more reliable ways to do this.