I think I'm understanding the idea you have about the carbon dioxide container sandwiched between your phase-change cooler and the CPU. First off, I'm not saying option 3) that you're a retard. You can have ideas that don't work and still not be a retard. It's just an idea.
Which leads me to my next point: the idea won't work. Think of the energy balance. The amount of heat going into the container of CO2 must equal the amount of heat coming out of it plus the heat consumed (or - heat produced) inside. The only way the CO2 will consume or produce heat is if the volume of the container is changing (for all intensive purposes it's not), or if the CO2 is undergoing a chemical reaction (not) or if the CO2 is changing state. If it's changing state (it starts as dry ice and sublimates, for instance), it can temporarily absorb some heat, but once the solid all turns to gas you're back to the old energy balance of what goes in must come out. It won't reduce the amount of work the phase-change cooler must do unless you're constantly swapping in new cartridges of dry ice.
As for other points, CO2 can be a liquid. This is studied in high-school physics, at least where I come from. As the OP said, it requires pressurisation. Look up "triple-point" and I believe you'll find what I'm talking about. At atmospheric pressure, CO2 goes straight from solid -> gas. At very high pressures, it will go through all three states. It gets me when posters ridicule others' supposedly idiotic science when the others were correct to begin with.