If you're trying to get 500 MPG, you can't afford to lose valuable energy to incomplete combustion. Lean burn will greatly reduce HC and CO emissions (until it goes so lean that it doesn't ignite, which is obviously undesirable in every way). Water injection takes care of the NOx.
http://en.wikipedia.org/wiki/Water_injection_%28engines%29
There is very little incomplete combustion in the modern PI engine at near stoichometric conditions, much better than you would find in any DISI engine. A lean burn, phi < 0.7, does not greatly reduce HC emissions, it increases them. And water injection does not "take care of" NOx. In DISI engine it often doesn't do a whole lot in terms of NOx emissions, unless you're operating at relatively high combustion temps, i.e. Zeldovich mechanism is dominant. Water injection often leads to greater inefficiencies as well.
water injection is also of interest because it can potentially decrease nitrogen oxide (NOx) emissions in exhaust.
How so?
And as for CO2, just getting to 500 MPG takes it really low.
That 500mpg is not an actual metric of the engine efficiency, its a measure of how far the vehicle can travel with an engine AND rechargeable batteries. The energy required to charge those batteries had to come from somewhere. Perhaps a coal fired power plant.
It is a legitimate technology and a good direction for cars to go. It pollutes less and reduces foreign oil dependency.
I'm not arguing that the development of hybrid technology, whether by an OEM or in someone's garage, should not be pursued. I'm just stating that some of the methods you are proposing and that other people have developed actually do more harm to the environment than good as well as the way people measure the performance of these methods are incorrect and/or misleading.