There's also the issue of money. Developers who create a program they'd like to sell will have to put it on a password-protected web page - after all, the program is the web page. And that opens up all sorts of security and accounting issues that make the matter more complicated. Even if it all worked, I don't see developers making a lot of money off of what are essentially nicely-formatted web pages.
Finally, there are limits to what can be done, programmatically, with Ajax and CSS. Those limits are high, of course, but I think something like a client for remote access to my home Mac (a program that exists today for my Treo) is going to be darn near impossible to pull off in the browser-only space that Apple has provided to developers.
For the real innovation to be seen, Apple must give developers access to the entire iPhone, not just its browser. I want to see what nifty useful tools developers can come up with when they can, for instance, create apps that use the multi-touch interface in new ways.
To paraphrase a couple of developers I overheard as they were leaving Monday's keynote: "They sat down and worked on this problem of developer access for six months, and all they could come up with was 'create a web page?' I'm not impressed."
While I doubt Apple has spent six months on the issue of developer access - the company has been much too busy working on finishing the phone itself, I'm guessing - I do agree with the conclusion many developers seem to be voicing at this week's conference: I am not impressed with Apple's current solution for giving software makers access to the iPhone.
Just a couple weeks ago, Steve Jobs had this to say about the topic of third party applications on the iPhone: "I think some time later this year we will find a way to let third parties write apps and still preserve security.... We would like to solve this problem, if you could be just a little more patient with us, I think everyone can get what they want."
If Jobs had been referring to Monday's announcement when he made that statement 12 days beforehand, he would hardly have used the phrase "later this year”. The implication here is that this browser-only model is a stop-gap solution, and that there's a more full-featured approach in the works for later this year. Let's hope so.