I would initiate a steady program of planetary probes, a common interplanetary communication system, and space-based observatories.
Budget: $4B/year for about 6 probes a year over 2006-2016
about 30 heavy and 30 medium launches
Budget: $8B for 3 observatories over 2009-2016.
about 10 heavy launches
Most probes are limited by their ability to transmit data back to Earth. That ability is limited by the power available to transmit. For the planets beyond Earth (or perhaps Mars), power is limited because solar cells do not have significant yield. I would fully fund the Prometheus project, which is developing a small nuclear reactor for use in space.
Beyond finishing the development of nuclear reactors in space, existing technology is already comfortably close to the limits of what can be transmitted for a given amount of power. It makes sense, then, to have multiple probes in close proximity use the same communication system. This is already done to some extent at Mars, with Mars Odyssey and Mars Global Surveyor relaying signals for ground rovers.
I would initiate a program to launch a nuclear-reactor-powered long-range communication satellites into orbit around each of Venus, Mars, Jupiter, Saturn, and Uranus, and (probably) one roving satellite in the asteroid belt. These satellites would primarily be responsible for relaying high data rate bit streams to earth from local probes. They would be built for a useful lifetime of 25 years. They would also carry a modest scientific payload.
Budget: Prometheus: $4B over 2006-2010
Budget: Comsats: $12B for 6 comsats over 2011-2016
about 20 heavy launches
Orbital Assembly Development
Observatories and comsats, and to some extent probes, have one thing in common: very large communication dishes or imaging reflectors. Observatories and especially the extraterrestrial comsats will also be very heavy. All of these missions can benefit from orbital assembly. I do not forsee intricate assembly of the sort requiring people, nor do I see the ISS as a good place to do this assembly, as it will likely involve close proximity to very large amounts of propellant. I would fund one assembly robot in low earth orbit.
Budget: $2B for one assembly robot over 2006-2010.
2 heavy launches
The extraterrestrial comsats in particular and long-range probes in general have large propellant requirements. To avoid the need for seldom-used massive launchers and their infrastructure, these probes should be boosted to their destinations by the same stage that put them into low earth orbit, but refueled in orbit by other rockets, so that multiple launches can assemble the mass to be launched out of LEO.
To help the nascent boost industry, NASA would be required to fit all payloads into the standardized EELV payload masses and sizes. Payloads intended to fit in the Shuttle for ISS delivery will fit into Heavy EELV boosters.
The Shuttle program would be summarily dumped. If ATK thinks that solid rocket boosters are a good match to LEO delivery (as they may well be), they are welcome to collaborate on an EELV booster to compete with the existing 2 (Delta 4 and Atlas 5) and potential 1 other (Falcon 9).
NASA would develop a Crew Transfer Vehicle. This vehicle would be a capsule carrying about six that would ride as one of the smaller EELV payloads (perhaps the 9000 kg class). If t/Space thinks they can drop-kick people to the ISS for less money, I'd entertain a proposal.
Budget: $2B over 2006-2008 to develop the CXV.
The International Space Station would be completed with EELV Heavy launches over the next several years.
Budget: $4B over 2006-2010 (just for cargo launches)
about 20 heavy launches (ISS segments cargo only)
As well as assembling the station, crews would be sent up fairly often to perform science experiments.
Budget: $1B/year over 2006-2016 (just for the launches)
5 medium launches per year.
There isn't going to be any of this in the next 20 years. Instead, NASA would spend some portion of it's budget learning how to live, work and do science in space. In the meantime, all the probes launched would find out what we actually want people to look at.
Budget: $2B/year over 2006-2016
My total yearly budget ends up about 70% of what NASA spends today. Probably I have no idea how much things cost. In particular, I left out ISS operational and future segment build costs.
I don't see any massive increase in the launch rate. Over the next
11 years, I see about 160 launches with a total price tag of about
$12B. A billion dollars a year seems like enough money to keep
perhaps two players going -- with just 1000 employees each.
Heavy: 82 @ $100M/each
Medium: 80 @ $ 40M/each