I don't know whether SpaceX has fitted their boosters with sea anchors yet, but I wouldn't be surprised to see this addition after the present loss. In windy conditions, a sea anchor provides some drag and keeps the boat (or booster) pointed perpendicular to the waves, thereby minimizing the force expended on the object.
> Surely sea conditions are something they'd take into account when planning and executing the mission.
Ocean experts are constantly preparing for extreme conditions, then discovering they didn't prepare for conditions that were extreme enough.
> Wouldn't you have to launch that into space, too?
Yes, which is definitely a drawback -- every ounce that isn't payload is regarded as a serious waste, because it all has to be launched at great cost (the ratio of payload to fuel is often greater than ten to one).
On the other hand, the SpaceX people report that the booster was destroyed, so this well-tested idea is very likely in their minds right now.
I'm sure there are safety concerns that have prevented this until now. The successful retro-file to zero velocity over water may help resolve those concerns.
Sea conditions were too rough for recovery. Their options were to wait for better weather, annoying the customer and possibly losing the launch (they had one more launch window after the one they used before the payload went to another launch company), or go for it and still accomplish a great deal of testing even though they probably couldn't recover the rocket. Total no-brainer to choose the second option.
It goes beyond NASA as well. If SpaceX can demonstrate that their reusability testing doesn't impact the launch, that increases the odds that their other customers will allow it on their launches too.
Their next attempt will be to target a smaller radius, to prove their aiming accuracy. Finally, after a couple more ocean landings (with possible recovery for analysis purposes), they will attempt to land on dry ground. To do this, they need to prove they can set down within a 1-mile radius. But they believe they should have the same precision of a helicopter.
After this, they are aiming for same-day reusability.
I'd expect maintaining pressure in the tanks not to be a significant problem, though it would need to be addressed -- pressure needs to be maintained reasonably constant during the whole burn, and it's done by admitting helium from compressed helium tanks. If the fuel/oxidizer tanks don't leak too much, the same mechanism could be used after the rocket has landed.
[1] http://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/2008001...