An new optical telescope based on this would be cheaper than spending an Orion mission on fixing Hubble: http://forum.nasaspaceflight.com/index.php?topic=29081.0
Cheapest option would be to simply, again, farm it out to the private sector to bid on. They either propose a robotic craft like Jim is proposing, a robotic craft with arms that can do a similar job as astronauts, or a trip in a Dragon, Dreamchaser, CST-100 or that Blue Origin capsule. Lots of options, all cheaper than using an Orion. ESPECIALLY if you use both an Orion and an SLS launch.
Hubble orbital reboost delta-v requirement most likely means servicing would be an Orion mission. The orbital reboost might be the most critical servicing issue.
Why do you presume that no other vehicle can provide additional delta-V for HST?
For vehicles that can carry a crew, the Orion with its Service Module offer a total delta-v of 1,595 m/s and both have some useful designed-in long-term mission capabilities.
After doing the servicing mission, and with an Orion no longer attached to the SM, a modified ESA Service Module could offer some significant delta-v possibilities for the Hubble Space Telescope. Other mission profiles could be used.
Orion's origins included some possible useful Orion Service Module payloads and diverse capabilities for LEO missions:
INPUT TO THE REVIEW OF U.S. HUMAN SPACE FLIGHT PLANS COMMITTEE LOCKHEED MARTIN CORPORATION ORION STATUS AND MULTI-MISSION CAPABILITIES"Currently, the program baseline is to qualify Orion for both a two propellant tank and a four propellant tank version. By flying only two of its four propellant tanks, Orion can use the additional volume to perform ISS resupply missions for un-pressurized cargo. Additionally, the Crew Module can deliver and return pressurized cargo depending on the number of crew aboard Orion. This cargo concept can deliver approximately 1,800 lbs of pressurized and unpressurized cargo every crewed mission (four crew), or approximately 3,600 lbs of cargo per year at little or no additional cost."
And, "Orion could also provide 100% of ISS annual cargo requirements at a cost equal to or less than the current commercial approaches to sup-plement planned commercial systems. This concept employs an upgraded autono-mous Orion Service Module as a tug once the primary ISS mission is complete and after the Crew Module has been delivered to its de-orbit target."
And, "In addition to providing cargo services in conjunction with Orion’s nominal ISS missions, the upgraded autonomous Service Module as described above could be launched to deliver 30,000 lbs of unpressurized cargo including additional nodes, large spares and Orbital Replacement Units (ORUs) if the ISS continues to operate beyond 2016."
And, "Orion can support NASA’s Science Mission Directorate by hosting payloads and delivering or servicing spacecraft. A remote manipulator system can be in-stalled in one of the cargo compartments and on-orbit servicing for the ISS and other spacecraft can be performed."
Using the ESA Service Module as a tug might eventually offer several interesting options for capturing cargo vehicles the are not ISS-qualified and taking those vehicles to the ISS or to another space station, but the main issue of this thread is using the Orion and, or, its ESA Service Module to do maintenance and reboost on the Hubble Space Telescope.
The important questions in terms of a crewed Hubble servicing option revolve around what will be the real capabilities and options that will be available with the Orion and ESA Service Module.
If reboost, electrical power, and three axis control of the Hubble Space Telescope are the main issues for a Hubble servicing mission, then perhaps only launching a suitably modified ESA Service Module that semi-permanently docks to the Hubble would be a suitable answer. No Orion would need to be flown.
If a "remote manipulator system can be in-stalled in one of the cargo compartments and on-orbit servicing for the ISS and other spacecraft can be performed" without the presence of a crew, then depending on what is needed, that option may be the best way to accomplish a Hubble servicing mission.
Perhaps if an appropriately modified Orion and ESA Service Module have some robots and a "remote manipulator system" they would be quite capable and could return the captured asteroid samples from a Lunar orbit without needing a crew on board.
Obviously, there might be no real need for a crew and Orion to travel to the Hubble if the ESA Service Module and some robots could do the servicing mission.
Edited.