(Text of what I submitted to a comment section on the Transocean website:)

For future spill incidents, your company might want to construct a mobile system specifically designed to control deepwater spills.

Start with a square framework of ballast tanks, reinforced at the perimeter for structural strength, and about a hundred yards on a side. Equip it with thrusters for positioning, and remote operating links. Across this framework put a large "tent" structure of reinforced material, strong enough for the job. If you somehow inflated this thing on land, it would look like a small domed stadium.

Position it, and flood the ballast tanks. If the area directly over the well is afire, or flooded with oil, it may be possible to submerge it at a distance and move it into position underwater. Guide it slowly down to land over the well. The "tent" will make this difficult, but take it slow. There may be some way to temporarily vent the "tent" for water flowthrough during the descent, but this isn't essential.

Have tubes from surface tankers connected to the "tent" to tap oil floating up into it. Foam floats attached across the surface of the tent exterior would maintain its shape if the oil floating up doesn't. You need to maintain room underneath for ROV work. Gaps in the frame/ballast tank perimeter would allow access by ROVs. The system could have large lights and cameras inside to illuminate the repair field under the "tent." Maybe the system would have its own ROVs parked in the frame.

When the emergency has passed, and the well is sealed, blow the tanks and bring it back to the surface to be docked and used in the next emergency.

The system could be developed in stages using subscale models and tests in shallow water accessible by divers and observers.

A major investment, but not compared to the feat of building deepwater rigs in the first place, and not compared to the money major spills will cost the oil companies.