I just saw this article discussing the challenges of returning from Mars.
Returning from Mars is a large engineering problem that's independent from most (not all) aspects of getting there, and it's a good idea to start pondering the problem even though a specific mission architecture hasn't been agreed on.
What's sad is that ATK, Lockheed, and Grumman were NASA's choices. This is the root of many of NASA's problems: an inability to divorce itself from the lumbering herbivores that have grown (over the last 5 decades) to define the agency. Corporate behemoths like the ones named here (and several others) are where good ideas and creative thinkers go to die. Tiny organizations have great ideas and accomplish amazing things with limited resources - this is the essence of engineering, and that type of thinking and motivation is what got us to the moon originally (yes, I realize that it was the same contractors back then. That was 40 years ago). Now, NASA meets its small business goals by having large contractors subcontract work to small companies.
How does this manifest itself? A couple of examples:
This is how the large contractors think, and they are consistently rewarded for these inefficiencies and the hundred other examples of syphilitic idiocy they dream up each day. One lesson presented by the Orion program (I no longer refer to "lessons learned" because it's clear that we learn nothing) is that the large contractors preserve their profit margins by re-using as much of their aging junk as possible, because innovation takes time and energy. Thus the "high-tech" glass cockpit of Orion was going to use LCD displays that were obsolete before PDR, because that's what the large contractor tasked with making them had lying around.
I'm sure that given enough time and money, this set of contractors could make a functional crew return vehicle for a Mars mission. But what are the odds that they can make an excellent crew return vehicle?