You can't run these things inside of a browser?
http://games.yahoo.com/games/front
The writer(s) of Pickleview have already demonstated that you can make a darn complelling app run inside of the browser.
http://pickleview.com/
I don't think the issue is running them in a browser, or that compelling apps can't be run on the browser, as that has already been proven possible. The issue I believe is that you are creating an additional hurdle/system requirement on these apps that is from a purely programmatic standpoint utterly unnecessary.
First consideration.....I'd LOVE to be able to run some of the available apps when on a plane, when on a train into the city, and lots of other areas where I do not have either phone or EDGE access. AFAIAC conditional access is a HUGE negative mark on considering the quality of an application. I mean, it's like if I couldn't use the iPod software, or the Video software, while I don't have EDGE access. If the app is such that its operation involves network access (for example, IIRC someone wrote a webb app for the iPhone that does remote server monitoring) that is understandable. However, if it does not...
For example, I tried to run a Bejewelled-type software on my iPhone. This involves opening the browser, hitting the link, waiting anywhere from 2-5 minutes (depending on edge quality) of downloading to play the game. My friend who has modded his iPhone hits the icon for Lights Out! and in the time it takes me to open up Safari he's already playing. To abstract....if you built a really slick, well designed, well executed, functional, and useful piece of software, then sold it on the market, but the way it was designed every time a person wanted to run the progrem they'd have to go to the internet, log on, download the whole application, and then run it.....you wouldn't be making a lot of sales.
It reminds me of...I think it was the mid 90's....when everyone was all abuzz about web apps. They were the future....computers would be sold with ultra limited OSes and all the apps they needed would be in repositories they'd just access through the network. People imagined worlds freed of costly software updates, and storage, and system configuration issues, as all that would be handled on these mystical repositories. You'd just pay a small licensing fee, log on, and run all your office apps, email apps, and development apps from there. Thin cients and happiness for all! Didn't happen because it was wishful thinking and absolutely impractical for a dozen different reasons. I don't see anything about the iPhone that changes them either.
While I believe that there will be hundreds if not thousands of apps written for the iPhone I don't think they will be that successful. Possibly something totally new will catch on like the spreadsheet did for the PC.
It already has, in my opinion. Installer.app. You load one piece of software on your phone, and not only does it give you access to what at this point might be well over one hundred different apps, it is an amazingly streamlined content delivery system. Tap the app, tap install, and it installs the app for you. It even keeps track of your installed apps and will notify you if an update exists, which is just as easy to apply as the main installation. A well designed and well executed concept, and all this done on a platform in which it was never designed to run. Considering all these factors, that NOTHING like that exists for cell phones, and the popularity of the software....I'd definitely call that a success. I'd also take my hats off to that level of programming skill. As to whether or not others will be "successful", that depends on how you define success.
Honestly, though....I don't think the venture is going to be that successful either, as I see Apple going very much the way they did with ringtones. They're going to license a select few developers (small time and single person dev houses need not apply....and forget rolling your own) who are going to develop a select few apps, in order for Apple to keep control of their system. You can only get them through iTunes, and they're gonna cost....my guess in the area of $5-10 bucks a pop (my guess being 10, as rumor has Apple looking to get a pretty big chunk of that scratch). Don't know what they going to do re: charging for updates, but regardless the vast majority of apps that peope currently enjoy on their hacked phones (things like Summerboard, IM clients, etc) absolutely will not pass Apple's muster and thus won't be offerend on the iPhone via their system. Plus I think it won't be till we're well in the '09 timeframe that Apple will even come close to offering the breadth and depth of software that one can obtain via Installer.
I'll borrow a quote from Ian Rogers at Yahoo from his Digital Medium Forum presentation: "Convenience wins, hubris loses". If the users had a choice between a web app and a native app, they'll choose the native one because its more convenient. It's why Apple is even entertaining the whole concept, well, that and it is another revenue maker for them. I just hope the extend the concept to their implementation, but owing to past experience I don't see that happening.