In sept-oct-nov Ad Astra, there's an interview with congressman Dave Weldon. He favors a capsule.

It's more affordable to develop and operate, is safer, has the most promises for non-LEO missions, has more robust abort modes in all phases of ascent.
A capsule design (symetrical truncated cone, cone, bell or sphere) has no concerns if enough lift is available at the time of emergency seperation from the stack, as would a winged vehicle. Wings, wheels have failure modes non existant for capsules.
No wing or wheels incrases the cago weight capability.
A modular upgradeable approach allows modules to be stretched and modified to handle different missions.
No concerns about weather at abort sites during launches.
A capsule will undergo a less complicated test program, which means less time and less money.
There would be an assembly line which would keep people at work; No more search at eBay for spare parts no longer manufactured.
No need to inspect the vehicle after each flight (if it's non re-usable).

About the Launch Escape System:

I wonder if it couldn't be made re-usable? With a parachute...
I also wonder at what moment it can be jettisoned (to save weight); once the rocket is at a certain height and speed and with a good momentum, even if the rocket exploded, the capsule could save itself just by being ejected from the blast of the explosion and the heat shield protecting from the blast.
Not a single Ariane rocket exploded at the moment of liftoff ;the first Ariane V exploded a few seconds after undergoing dynamic imbalance and this about 20 seconds (I don't remember) after liftoff. Pad explosion has become very rare, even for non man-rated ELV's. But of course, a LES will be a necessity; it's just that there can be ways to make it more efficient or less costly.

Raoul