Space is a complicated arena... unfortunately, politics plays FAR too big a role in it...
Someone mentioned that NASA is merely following orders... quite correct... if ordered, they will downselect to a single source provider. They have to answer to the Congress which controls their purse strings. BUT, the NASA Administrator and Deputy Administrator (Charlie Bolden and Lori Garver) are appointees of the President-- Lori Garver was his first choice, but alas she was deemed "unacceptable" to the Senate, who made no secret they wouldn't confirm her... Charlie Bolden, on the other hand, WAS considered sufficiently "acceptable" to the Senate to confirm his nomination to the position. BUT, make no mistake, this "commercially driven" approach to space is very much part of Obama's intended space policy, and it's surprisingly a good one... Congress, especially the House, is diametrically opposed to it because it would shake up the Space State constituencies and the political machine feeding them, via preferred NASA shuttle contractors, money going into specific NASA centers in various states, and other such government largesse to aerospace industry campaign contributors and lobbyists and their constituencies... Hence the House's VEHEMENT resistance to the 2010 budget proposal that killed outright, and the Senate compromise with the Administration which ultimately was adopted, that killed Ares I and Ares V and the Constellation program (and which nearly killed Orion as well, which hurriedly "morphed" into the MPCV under the compromise). This compromise ended up trading Ares I and V for the SLS, which was basically a return to the original plans for the Ares V before the so-called "1.5 launch solution" using Ares I really screwed up the works... BUT, with SLS not even going to be available for test flights until the end of the decade, (or more likely somewhat into the next) that, by default, leaves LEO crew launch ops to ISS up to either 1) the international partners (Russia) or 2) commercial providers in the US. Congress doesn't like the idea of commercial because they intrinsically know that commercial operations are going to be run SO much more efficiently that when they DO succeed (and no it won't be overnight, or easy, but they will eventually) that it's going to spell the end for the "gov't designed and contractor built" gov't/big aerospace contractor arrangement... which provides a lovely revolving door between gov't and industry, as well as all that nice lobbying money and such, influence peddling, campaign contributions, etc... it will upset the applecart. This is EXACTLY the same reason why, when after Columbia when the decision to retire shuttles and replace them with a CEV capsule and expendable launcher, when it became apparent the front-runners for the job were going to be the existing but modified EELV's, Congress was up in arms... and through some political tradeoffs managed to have removed then-NASA Administrator Sean O'keefe and Admiral Steidle, who came up with the "spiral development" concept of creating first a capsule, launching it existing rockets, adapting them into higher payload capacities, then developing payloads, and doing the entire program in small, achievable, affordable, simple steps building on one another over time toward the ultimate goal. Congress demanded a "shuttle derived" solution, to keep the existing "shuttle gravy train" of contractors and various NASA center pecking order pretty much intact to the extent possible, even if it wasn't affordable or desirable as a solution... Hence the appointment of Mike Griffin as NASA Administrator under Bush II, who immediately switched gears to an "all shuttle derived" solution for shuttle replacement... which became Constellation, which subsequently spun its wheels for over half a decade and ultimately foundered and was canceled early in Obama's Administration (and rightfully so, I might add).
SpaceX has done some wonderful work... sure they're just starting out, but they've delivered two new vehicle designs (Falcon 1 and Falcon 9), an unmanned cargo capsule (Dragon), new engines (Merlin I-C and the airlit Merlin upperstage vacuum engine) for less than the pre-development work that was done on Ares I and Orion... FAR less, in fact... They've proven that space CAN be cheaper if it's streamlined into an integrated organization with motivation to cut the gov't red tape and eliminate the contractor overbilling and delays as much as possible, which are hallmarks of the NASA/contractor cost-plus contract methods NASA has been using all these decades. SpaceX still has a lot to prove, and I don't want to be accused of being a fanboy, because I'm not... but they have a good record so far and I certainly hope their success continues!
Orbital has done some good work, though their success is marred by incidents such as payload shrouds on their launch vehicles failing to separate, dooming at least two NASA payloads. They're also further behind than SpaceX, who has demonstrated their rocket and Dragon capsule in an actual orbital flight and recovery. Orbital is probably at least that much further behind on any contemplation of manning their Cygnus ISS COTS resupply vehicle...
ULA's CST-100, borrowing heavily from the Orion/MPCV design, should be a fairly straightforward contender. ULA also has the advantage of a well-proven launch vehicle in their Atlas V rocket, which is to be the launch vehicle for the CST-100. Modifications of the Atlas V should be pretty straightforward since the engines, the Russian RD-180's, were originally designed for use on a manned vehicle (albeit in a different configuration). ULA would probably be the front-runner head-to-head against SpaceX, but SpaceX also brings the advantage of fresh ideas, having not been operated for decades as a government cost-plus contractor feeding at the gov't trough...
Dreamchaser is an interesting shuttle-like crew vehicle, but it hasn't flown yet and is still years away from flying. It too would fly on the Atlas V. While an interesting idea and probably practical if proven that it works reliably, it's also riskier and probably further out than the other contenders, IMHO.
ATK's cobbled-together entry in CCDev (Commercial Crew Development) hasn't gained much traction, being a cheap retread of the Ares I design, cobbling an Ariane V core vehicle liquid stage on top of the five-segment Ares I solid rocket motor first stage from Ares I. ATK recently trotted out their capsule design, based on composite pressure vessel designs they constructed prototypes of and studied for NASA as possible weight-saving substitutes for the aluminum-lithium pressure vessel of Orion. Having a basic pressure vessel is a LONG way from having an actual functioning spacecraft... it's like the difference between having the basic stripped body shell of a car sitting in the garage and having the engine, transmission, electrical systems, engine control computers, chassis electronics, and all the supporting suspension and drivetrain components installed and working reliably... or comparing a bare engine block to a fully complete, tuned, and balanced racing engine that's been dyno tested. I don't put much stock in ATK's ability to make this kludge work, but I don't have a problem with them trying... who knows, they might surprise everybody... stranger things have happened.
This whole mess is basically Congress trying to "manipulate the process" by injecting themselves in such a way so they can leverage the outcome for political gain. The space program is SUPPOSED to be about exploring space-- not about political clout and kickback here on Earth... BUT, unfortunately, NASA was born into such an atmosphere and operated within it all these decades, and probably will die in such an atmosphere...
When/if commercial truly succeeds, it will be a HUGE blow to the gov't control of aerospace and the need for lobbying and kickbacks and stuff that keeps the space money flowing into the various preselected states and companies... and it will immediately call into question the need for large, expensive, multi-decade programs like SLS to build HLV boosters under the old cost-plus NASA design/contractor built government controlled system, with all the political benefits that entails. Reducing NASA to the role of advisor and seed money provider and ultimately paying customer rather than design and contract awarding controlling agency at the behest of the federal government would jeopardize this lucrative arrangement that exists now under cost-plus, both to the political players on the Hill and to the big existing aerospace contractors.
Someone mentioned, "well, if they're commercial, let them develop it and NASA can buy it... why the gov't money?" Simply because space systems are not new cars-- the market is very small (even for unmanned commercial launchers; for manned systems its an order of magnitude smaller!) and the systems are SO expensive to design and construct that companies have been historically unable to get sufficient investment capital to afford the design and development phase... in short, there are MUCH less risky and more highly profitable investment opportunities in much shorter time spans than investing in space booster or spacecraft development, hence the money flows into those "sure things" rather than risky space ventures that won't pay off for years or decades if ever. It's much like the military's means of procurement... the Air Force (or whatever service) doesn't design their new bomber or tank or weapon system or whatever, they merely assess the needs and desires for the new system and how it would work and integrate with the existing weapon systems, transportation, logistics, operations, weapon systems, ancillary systems, etc. and then issue the requirements, and put it out to bid in industry. They then receive the bids from industry as proposals, evaluate them, and select a few for design and development. These prototypes are then tested and the "winner" selected for production. Sure, there's some "preferences" that come into play, political considerations and such, that influence the choices made of who the winner turns out to be... but it's a more transparent process that produces better results than the "NASA designs it, contractors build it under cost-plus contracts" outmoded model that NASA uses, that hearkens back to the beginning of the space and missile age when NOBODY really knew how to design or build a missile or rocket and its systems, but a select few experts, and as such most of the design and development HAD to be done by this small group, who then turned the designs over to aerospace contractors to build... The gov't pays for these prototypes via development contracts, because the sheer costs of the systems design and prototype construction is beyond the means of most industry players to pay... especially for a system that will, for quite some time, probably only have ONE single customer... Uncle Sam, and perhaps a few more later on to selected Allies via arms sales abroad to foriegn governments.
It's HIGH time for NASA to get out of the design and rocket building business and simply issue requirements to industry, for them to bid and submit their designs to NASA for scrutiny. NASA can then assess the merits of the designs and downselect for further development or award the contract for a sole prototype system development at that time... similar to identical to how the Armed Services obtain their new systems... THAT is what Congress is trying to stop...
later! OL JR