The Set America Free folks have the right idea. Their proposal is a series of legislative steps that we can take now that will have the effect of reducing our oil imports through a combination of better fuel efficiency and generation of oil substitutes.
I'm not 100% in agreement with these folks. They'd still fund hydrogen fuel cell research to the tune of $2 billion over the next 4 years, and they pay lip service to biodiesel research. I'd whack that fuel cell research completely, and put steady money into biodiesel for at least two decades -- $300M/year into biodiesel from crops (this will help work out the bugs in delivering biodiesel through our supply chain and any vehicle use issues), $100M/year into biodiesel from existing microalgae (to develop and debug the infrastructure for growing and processing for millions of tons of algae), and maybe $100M/year to $200M/year into engineered microalgae.
The oil business is so huge, and leverages such a massive prehistoric biological mechanism, that replacing our oil imports will eventually involve re-engineering our environment in the same way that we have done with building big dams, draining big swamps, and farming the prairie. We are never going to produce meaningful amounts of fuel from farm crops, because there is not enough land, not enough fresh water, and land plants produce too much non-fuel mass to sort through. When Big Oil gets into biodiesel, we'll end up extending Louisiana, Texas, and Florida with dikes extending into the shallows of the Gulf of Mexico. Within those dikes will be massive brackish algae ponds pumped to saturation with the CO2 from coal-fired powerplants. Millions of gallons of oil every day will be extracted from algae separated from those ponds. Environmentalists will be outraged at the devastation wrought on the delicate marine environment. Folks living on those coastlines will protest the change in their views (but will probably welcome the new jobs).
And negotiations with fundamentalist regimes in the mideast who wish to have nuclear weapons will have a decidedly different tone. Without the U.S. dependent on Iranian and Saudi oil, without their ability to tweak our economy with simple price changes, without the constant flow of Westerners who must tend to their oil fields, without the massive influx of cash that keeps their corrupt governments in power and able to support expensive research into WMDs... things will be vastly different. Conditions will be more like Afghanistan and less like Saudi Arabia. More like Afghanistan but without the subsidized madrasas. But that is their problem, not ours.
But none of this is possible now because nobody really knows how to grow a lot of algae cheaply, just like nobody knew how to use steam and CO2 injection to profitably leach oil out of a recalcitrant well 100 years ago. It is a U.S. national security priority to develop extensive domestic energy supplies, and that is why the U.S. government should fund biodiesel research. We had a decent program going for about a dozen years, called the "Aquatic Species Program", started during Reagan's tenure, which was canned by the Clinton administration. Here's their report. (300+ pages, check the table of contents and the last couple of sections for the good stuff.)
So there is a lot of research and development to do.
I should say that by engineered microalgae, I mean crop development the way it's been traditionally done for thousands of years: grow a lot of algae, select the stuff that produces the most oil, propagate that strain and wipe out the rest. Iterate hundreds or thousands of times. Set up multiple centers across the U.S., so that various strains can be developed independently, optimized for the local environment, adapted to the highly acidic environment we want to grow this stuff in, with different approaches to optimization by different teams. Given the rate at which microalgae grow, I think we should be able to get the iteration time down to a week at maximum. One iteration a day would be significantly better. I'd expect some decent results within a decade. If illiterate people can take wild maize and transform it into corn in a millenium with just one or two crops a year, we can turn wild microalgae into a serious oil producer in a decade.
I wouldn't hold out much hope for actual genetic engineering of the microalgae. Genetic engineering is good at turning off particular pathways inside cells. It might be useful for adding pathways that wouldn't exist otherwise, say, if you want to produce a particular drug in carefully sterilized bacterial fermenters. But the problem is that any tinkering we do is going to make the resulting species less well adapted to its environment. Algae in the wild live in a very competitive environment. Algae farms are going to be cheap places, not well-controlled places -- we probably can't afford to even put thin plastic film over the ponds to cut down on evaporation and CO2 loss. So we can't protect specially engineered algae from competitors.
I think there is a lot we can do in the next decade to turn around our crummy national security situation and maybe improve our economy as well. We need to get Congress on board, define the problem, and eliminate distractions.