I'm sitting with my iPhone in one hand, reading your post, and before this I checked my Facebook account using the native iPhone client and now I'm wondering ... what the hell are you talking about?
Do you think the hard part of building something like Facebook is that stupid native client (that can sync just fine)?
Do you think we make our lives easier by only making the "hard part" of a job easier?
Look at the number of little webapps people here make. Do you see absolutely no value in also being able to produce native apps at the same time, with dramatically less additional work, using the tools and languages you already know?
Facebook's native client "syncs just fine" because they did extra work to make it so. Why is it bad wanting to get that functionality for free?
I understand your point mostly, but I think you're wrong nonetheless.
They did get that functionality for free (mostly) with their web client optimized for mobile: which works and behaves just fine. But even so, they rolled their native client anyway, because the web client doesn't feel and behave like a native iPhone app.
A native iPhone app is property optimized for the screen width, it does local caching (which a web app used on a desktop doesn't need to do), it behaves differently in regards to the control used (which respond to different events), and in many cases it is desirable to get rid of the page metaphor, no request/response cycle, etc...
The business logic may be the same, but the UI logic is very different and you don't want something autogenerated because the UI itself is usually a competitive advantage; and assembling it is not the hard part, a lot harder is to come up with a good design or to make your backend stable (I don't know how it happens but when I go to Foursquare 6 times out of 10 their service is down).
For instance ... I'm using Twitter a whole lot more on my iPhone because the native client kicks ass. In contrast I use Facebook more on my desktop because the Twitter's web interface sucks (yeah, I'm weird like that).
What I'm trying to say is ... companies that want cheap transition to new platforms like the mobile, already have the right tools: the web and its related protocols; which are more ubiquitous than ever. If you want more than that (say, to differentiate yourself from the competition, or just to keep up), autogenerating the UI won't help; you'll have to get down and dirty because the beauty of a good design lies in the details.
I think you're missing the point. The UI isn't autogenerated -- it's the only bit that isn't. (Aside from DirectToWeb, which is a whole different thing). What WO allows you to do is save on all the backend work and focus entirely on the details of the UI. Which as you say, are the bits that matter.
There may be no silver bullets, but there is having a gun when everyone else is using three different types of hammer.
Do you think the hard part of building something like Facebook is that stupid native client (that can sync just fine)?