You may not be a rocket scientist, but I am, so let me clarify a few things here.

You seem to be confusing ULA and NASA launch efforts here. The Atlas V and Delta IV EELVs are commercial designs. Titan is retired, never to be launched again, and the future (and ultimate feasibility) of Ares I or V remains uncertain. Also, under point 6, Atlas and Atlas-Centaur are the same thing. Atlas refers to the first stage booster and Centaur is the second stage.

Drastically increasing the launch schedule of EELVs would be a tall order, necessitating a great deal of infrastructure development. Where all the money for this, and all these extra payloads you'd like to launch, will come from I have no idea. Right now the gov't is up to its eyeballs in debt, and is rapidly increasing that debt bailing out automakers, banks, and lining congressional districts with cash for votes. I'd love you see the increase in launches just as much as you would (it'd keep me employed), but it's certainly not realistic.

But I have to take issue with the basic premise that seemingly underlies your post here, which is that NASA (or the gov't in general) needs to be the one designing, building, and launching these rockets. Why? Why limit the launch industry to one or two designs with the NASA-approved stamp on them? (Which may or may not be the best vehicles for putting things and/or people into orbit.) Why not promote competition and increase the demand for vehicles in the launch industry by getting NASA out of the launch business altogether. Make NASA a purchaser of rides, not a supplier. The launch industry can then build and fly the designs it wants and let a multitude of designs compete. My dream would be to see another few Space-Xs pop up in the next few years. Thankfully we're actually starting to see a little bit of what I want with the ISS resupply contracts to Space-X and Orbital. I would be even happier, though, if NASA were out of the launch business altogether.

Here's what it would take to capture Apophis:

I'm assuming that the capture will take place on the 2029 encounter and that the perigee of any capture orbit is equal to the closest approach distance (this makes the calculations simple enough that I felt like doing them to post on slashdot).

So to capture into a circular orbit would take an instantaneous delta-v of: 4.18455 km/s. (Capturing into a highly elliptic orbit (e=0.9) doesn't reduce the delta-v by all that much (4.13677 km/s instead of 4.18).

Assuming the mass on Wikipedia is right (2.7e10 kg), and using space shuttle main engines (Isp_vacuum = 450 sec), it would take 1.65e10 kg of propellant to achieve the 4.18 km/s delta-v. So, a lot. (More than 1 million Delta IV heavies worth of payload capacity to orbit).

Going to a higher specific impulse form of propulsion, like Ion thrusters with Isp from 3000 to 30000 would take between 3.58e9 kg and 3.81e8 kg of propellant, respectively. Though Ion thrusters are about as far away from applying an instantaneous delta-v as you can get (except maybe solar sails or other neat little things like that).

In conclusion, it would take a hell of a lot of effort to capture Apophis.

This is exactly my point. You say that the station derives from an old study, and that much of the hardware used is old. Well, any Mars craft could be a blank sheet design taking the lessons learned from ISS and putting them to good use. And as for delays due to the shuttle, you again make my point. If NASA designs another heavy lift vehicle, it will be the only way to get a Mars craft (or pieces thereof) into orbit. What happens if that launch system goes down to a failure? You have the exact same situation you had with the shuttle. However, if you design your Mars craft from the beginning to use existing medium lift, you'll have multiple options to get stuff into orbit (Atlas V, Delta IV, Falcon 9, Ariane, etc).

Finally, while being able to put everything together as one big piece and launch it may simplify some aspects of the design, if we're going to really do worthwhile things in space (colonies, stations, mining) there will have to be piecemeal launches. Habitats and the like will simply need to be too big to be able to launch on a single rocket. Why not start getting really good at putting stuff together in orbit (or on the lunar surface) now? That way, when the expertise is really necessary, we'll have it. We won't have to stop and spend money to develop it.

I still don't understand the seeming obsession with heavy lift. Why develop and fly a new huge expensive rocket, putting all your payload eggs in one basket, rather than use a greater number of smaller, cheaper, existing rockets? The more rockets you fly, the more you have to build, and you can begin to take advantage of economies of scale and reduce the dollars per kg cost to orbit. Another advantage is that if your rocket does encounter some calamity, you don't lose your entire (much more expensive than the rocket itself) payload, but rather just a piece of it. Yes, flying your moon/mars/where-ever spacecraft into orbit a piece at a time means that you have to assemble it once you are up there, but that just puts into use all this lovely experience gained building the ISS. So, more light to medium lift: give it a chance.