about the only question I might be asking would be how much are they spending to develop this technology and how much money do they think they'll save--in short, what is the ROI for something like this?
Some hard numbers to throw around on something like this:
The current "price" of a Falcon 9 is about $70 million USD. This is how much you would be asked to pay for a standard Falcon 9 if you made a serious inquiry to sales @ spacex.com for a real quote, but that comes before special one time engineering charges or extra features and special handling. That price gets you a rocket, the basic range fees for an ordinary simple LEO payload delivery, and engineering data for a standard SpaceX payload connector. Most payloads usually require some special engineering considerations, so the price usually goes up from there.
The "cost" of a Falcon 9, in terms of raw labor and manufacturing costs to get that vehicle actually manufactured is definitely less than that figure, where I've heard the price is speculated to possibly in the very roughly $20-$30 million range to build that rocket, with most of that cost concentrated in the construction of the lower (1st) stage of the Falcon 9.... let's say about $15-$20 million for actual labor & manufacturing costs of the lower stage. That is a huge profit margin, which one of the reasons why Steve Jurvetson (a member of the SpaceX board of directors) is publicly quoted as saying the SpaceX financials are "financial porn".
Elon Musk also suggested that the fuel costs for a typical Falcon 9 flight are well less than $1 million per flight, more along the line of about $250,000 per flight to give a general ball park figure.
You can use that as a range of figures to try and figure out what the ROI of performing multiple failures and how many times you need to recover the lower stage before it becomes profitable. Also of note, this particular launch of the SES-9 vehicle likely would have been to purchase the full vehicle and not really involve any bonus for lower stage recovery where the expectation is that the stage would not have been recovered. It should also be of significant note that purely for R&D purposes that have nothing to do with reflying the lower stage, obtaining the physical equipment for engineering review is incredibly valuable where actual rocket engines used to delivery payloads into space and then put on a test stand for additional performance testing can help to significantly improve reliability. Just using a borescope to peek inside of the engine parts to see how they held up under actual flight conditions is alone worth the price of that stage if it means fewer disasters like the CRS-7 flight.
In short, I think you could likely make a case that even recovering one in three or more likely every other launch for that lower stage would more than pay for this whole recovery program effort from a strict cost accounting basis. I'd love to see what the engineering costs for this recovery development have been,but compared to the costs of developing the Merlin engines or the costs of developing the Dragon spacecraft, I suspect it is minor and almost incidental. A single successful recovery is worth at a minimum of $10-$15 million cash in hand, and that even includes the costs of the barge operations.
To show where SpaceX is looking from their own caculations, they intend to drop the price of the Falcon 9 to about $30 million with the regular recovery of the Falcon 9 lower stage, and if they can ever get the upper stage to the point of being recovered as well, that price goes down to a mere $7 million per launch that they intend to charge their customers. SpaceX has also announced their intention to perform payload faring recovery and reuse, just to show the extent of their reuse plans.