We didn't support 3GS because creating 2X the artwork would have been a lot of work for a relatively small % of iPhone users that will soon be upgrading to retina. Consider your comment a +1 for 3GS.
iOS app dev these days is all about increasing what I like to term the ARPH (average revenue per hour), a function of the revenue generated per hour spent working on the app. The best strategy IMHO is to get something out there ASAP, validate the market, and then expand if the data suggests a viable product that people will pay for.
Well we didn't want to do a simple scale-down, so some assets would have to be touched up, and other stuff would need to be remade. We're a small team so we did a cost/benefit and decided to focus on iPhone 4 and above.
