See the 2:40 mark of the video in the first post. The lander, which clearly masses considerably more than the Apollo Lunar Module, is deploying not one but 3 parachutes. These deploy after the back aeroshell is left behind, allowing the parachutes to unfurl. As the parachutes unfurl the descent engines kick in. I believe Steven Pietrobon mentioned that this approach chops the descent delta-v required from the engines to a mere 500 m/s. That's an impressively low figure for landing something 50 mt or more on Mars. I don't know what the figure would be doing an all-propulsive approach, but it'll be significantly more than that.
PAge 19 of this document talks about this.
Mars Exploration Entry, Descent and Landing Challenges (paper):http://www.ssdl.gatech.edu/papers/conferencePapers/IEEE-2006-0076.pdf
"Supersonic propulsive descent
Following hypersonic entry, a vehicle intending to land on
the surface of Mars must slow itself from supersonic
velocities to a speed appropriate for a soft landing. This last
deceleration phase, which involves only a few percent of the
vehicle’s remaining kinetic energy, has been initiated in past
robotic missions below Mach 2.1 using some combination
of parachutes and rocket-propelled descent. From Figure 20,
it is clear that a Mach 2 initiation of this phase is not
sufficient for the high mass entry systems associated with
human exploration. The total descent time from Mach 3 or 4
to landing is on the order of two minutes. During this phase,
several vehicle configuration changes are required. In a
matter of seconds, the vehicle will need to re-orient itself, an
aeroshell and/or back shell may be jettisoned, parachutes
may deploy, engines may start, navigation and hazard
avoidance sensors must operate, and landing gear may
deploy. In this very dynamic phase of flight, robust event
sequencing and timeline margin are critically important.
To date all parachutes utilized in the robotic Mars
exploration program have been derived from the technology
effort that led to the Viking flight project. These systems
have been limited to diameters on the order of 10-20 m and
supersonic deployments below Mach 2.1. As discussed in
Section 5, in an effort to improve landed mass, the robotic
exploration program may pursue a large diameter supersonic
parachute, likely no larger than 30 meters and deployed at
velocities below Mach 2.7 (in response to thermal
constraints). As a result of the large masses involved,
parachutes sized for human exploration systems would
represent a significant departure (in both size and
deployment Mach number) from their robotic counterparts.
In addition, due to their size, such systems will require
significant opening times. For example, to decelerate a 100 t
vehicle from Mach 3 conditions to 50 m/s near the Mars
surface would require a supersonic parachute diameter on
the order of 130 m. Similarly, a 50 t vehicle requires a
supersonic parachute diameter on the order of 90 m. While
clustered supersonic chutes are an option, the size of such
systems would still result in large timeline penalties for
opening. As such, an all parachute approach for Mars
human exploration vehicles, similar to the concepts now
used for robotic landers, is likely impractical."
It was something like this that I was thinking of. I guess it's not so much the parachutes can't be made big enough, but that they might take too long to deploy.
It continues with some more info:
"While parachutes alone are inadequate for slowing large
payloads at Mars, the all-propulsive solution results in high
propellant mass fractions and requires aeroshell separation
and propulsive descent initiation to take place at supersonic
speeds. As such, a trade study was conducted to quantify
how a large, supersonic parachute could mitigate these
issues. In this assessment, aggressive assumptions were
made in regard to parachute deployment conditions
(Mach 3) and altitude requirements for the subsequent
descent and landing events. Figure 24 shows the parachute
sizes required to decelerate a payload from Mach 3 to Mach
0.8 at an altitude of 2 km. A Mach number of 0.8 was
chosen to mitigate the aeroshell separation and re-contact
concerns of current robotic landers. Figure 24 shows that a
30 m, Mach 3 parachute allows for a subsonic propulsive
deceleration maneuver if entry masses are below
approximately 33 t. This same parachute can slow the
vehicle to Mach 1.0 at 2 km for entry masses less than 50 t.
For entry masses above 50 t, a larger chute is required (with
a significant opening time penalty), or the propulsive
deceleration maneuver must begin supersonically.
An additional benefit of this approach is that the parachute
can be used to separate the payload from the aeroshell.
Atmospheric uncertainty is a major driver for parachute assisted
descent. The results described above are for a
nominal atmosphere. If a conservative density is modeled,
the 30 m parachute is only practical for entry masses below
approximately 20 t. Parachute assisted propulsive descent
still requires significant propellant mass fraction to bring the
vehicle from Mach 0.8 to a soft landing. The propellant
mass fraction required for just the cross range maneuver (to
protect pre-landed assets on the surface) will actually
increase for a parachute-assisted system because the burn is
started much later in the descent. Overall, the total
propellant mass fraction required for descent and landing
will decrease from 20-30% of entry mass for an all propulsive
system (see Fig. 23), to a range of 12-18% for a
parachute assisted system.'
However, where this whole equation changes, is they are trying to maximize landed mass, understandably. And this lander is probably not trying to get back of the ground. However, any vehicle that needs to get back to Mars orbit, or do a directly return from the surface, will need a LOT of propellant. with large tanks for that propellant. So if you are designing to maximize usage of components and minimize elements rather than maximize landed mass, then it could be ok that there's a large dV penalty for propulsive supersonic retropropulsion. The lander would have the fuel tank capacity as it needs much more than that to get back to Earth. It lands burning up it's stored methalox, and then refuels on the surface with it's LH2 store. Then lifts itself back off the surface. And burns for Earth. Or rendezvous with a MTV and is discarded.
Such a lander would probably not be able to land very much cargo on the surface, but if it could actually work, it would maximize element use, and minimize individual elements. A simple system with minimal elements, but not perhaps very efficient in the amount of equipment it can drop off on the surface.
In a single new vehicle development, you have your Mars Transit Hab, your lander, your Mars Ascent Vehicle, and your Earth Return Vehicle. And you get to land it all back on Earth. The side advantage is this can probably take off form Mars and land on Earth with a decent amount of Mars samples compared to a more traditional architecture that only lands say, an Orion Capsule back on Earth.