Re: floating back to terra firma using three parachutes
Musk's landing by rocket is not about the position as much about wanting to reuse as much of the rocket and engines as possible. It makes great TV, not so sure about the science and economics given the extra weight and complexity.
It has everything to do with position. As his attempts to catch fairings demonstrate, the accuracy available with parachutes is relatively poor.
If you want to reuse your rocket then it needs to not go for a dip in the ocean, which requires that you be able to land it on a feasibly-sized barge (or in case of Return-to-Landing-Site, with a level of confidence that NASA/FAA will accept - they don't want a booster plopping down in Orlando or even the KSC Visitor Centre because that's where the wind carried the chutes).
Reuse hinges entirely upon the ability to accurately land on a pad - much the same as a 747 wouldn't get reused if you just cut the engines in the general vicinity of it's destination and parachuted it down.
As for economics? They speak for themselves. Musk has the cheapest orbital launch system on the market, and he's turning a profit on his launch business (SpaceX as a whole makes a loss, but that's down to R&D expense. The commercial launch business unit is structurally profitable - bankrolling R&D).
This doesn't apply to the capsule because it's a much simpler component and you can afford to replace a few salt-water damaged exterior components - but the expensive bit of the booster is the engines - there's no point in trying to reuse a F9 if you're going to have to replace salt-damaged Merlins. You'd be better off saving them and bolting them to a new booster body - which is what ULA have spitballed for Vulcan - jettison the engines, parachute them down, catch them in midair with a helicopter and use them on another rocket. You lose the rest of the booster which is really just a big (relatively cheap) empty fuel tank at that stage.