Sure, a cross-platform dev system might only support a common-denominator group of features. But equally it might add its own implementation of software features which aren't supported by a particular platform (e.g. widget sets providing "missing" widgets), or it might add ways of simulating missing hardware features from other platforms (e.g. the left-and-right-button-together emulation of the X-Windows middle mouse button).
Now if he was honest, he would have said straight out: "We require anyone coding for the iPhone to tightly tie their code to the iPhone API, so that porting to any other platform takes impractical amounts of effort." This might not be a popular statement, but it'd be true, and I think it'd accurately reflect their strategy.
There's a great historical precedent for this. The survival of Apple in the late 80s and early 90s can be summed up in three words: PageMaker, PhotoShop and Quark. The sole reason for using an Apple back then was DTP - a tiny minority of home users whose day-job was DTP also used Apples, but the vast majority of Macs were sold to DTP places. If any of those DTP packages had been successfully ported to Windows, Apple would have dried up and blown away like all the other dead platforms of the time (Acorn Archimedes, for example) which had their lunch eaten by Wintel. But porting software was hard enough that it was another 10 years before any of those made a half-decent move to Windows, and by that time Apple had their feet firmly under the table.
So far, it doesn't look like there's a killer app for the iPhone and iPad. There's a number of things it does fairly well, but nothing that really sets it aside from any other smartphone, and frankly I don't really see it happening. But my guess is that Jobs is thinking if a killer app *does* turn up, achieving some kind of lock-in for that killer app on his platform will let him completely stomp the opposition. As it stands the iPhone has first-mover advantage, but this doesn't necessarily give it the leverage to force everyone else out.
The downside of this is that by having a monopoly on who gets to sell their apps on your platform, you're leaving yourself very vulnerable legally. Sure MS only got a legal slap on the wrist with MS vs. Netscape, but it cost them a fortune in lawyers and some fairly hefty fines (particularly from the EU, which isn't politically dependent on a US company). And MS's downfall was failure to release details of secret Windows hooks that improved performance - they didn't completely block Windows PCs from installing Netscape, which is an extreme case of restriction of trade. Sooner or later Apple are going to piss off someone with sufficiently deep pockets (Google springs immediately to mind, but Adobe could equally well be it), or a smaller someone with sufficiently good political connections, who'll drag them all the way through the courts. If the MS vs. Netscape case is anything to go by, this won't have a massive effect on profits, but it'll do very serious damage indeed to the company's image.