Now, you could spend the same time making an app that you sell for $0.99 but you would need to sell about a thousand of them just to break even (since Apple takes such a huge cut, plus it costs more money to service a thousand people than it does to serve one). Furthermore your cash flow is much more predictable under a SaaS model (hence, easier to raise on, exit with, hire on, or just plain sleep at night with) so it isn't even a 1:1000 ratio. More like 1:2000.
The psychology behind this is pretty simple. Most business owners know that saving time increases their billable hours. If you can do that, then great. They would be thrilled to hear that they will be paying you that 12 bucks for the next 4 years, because it would mean that they survived at least that long! Most of them think there is a good chance that they will be toast in their first 3 or 4 months, and what's $40 bucks compared to their other expenses?
Whereas a general app purchaser isn't costing in their saved time (for business or productivity apps), since they don't view their mobile device as a work tool, and consumers don't usually care that your game is good because their are plenty of alternatives to choose from when it comes to wasting time on the internet.
Where I would build an app:
For an existing product that exists in a browser, mostly as a sales gen tool. Possibly for functionality that people would like to be mobile but doesn't work in a traditional SaaS app (check-ins for a ticketing app, for example, could be done on an iPad).
Most other cases I just wouldn't bother.
[1] I'd say 1.5 to 2% is basically the floor because that is roughly what small businesses close at.