There's more going on here. The orginal iphone had no sdk. So the only way to run stuff on it that wasn't built by Apple was via Safari. It ran ajax applications just fine. I actually worked at Nokia Research at the time and a few of my colleagues were responsible for porting webkit to Symbian, that was around 2005; around the time Ajax applications were becoming a thing before the iphone was more than a rumor.
For a while, there were many iphone optimized mobile websites. E.g. Google reader had an iphone version. Typically these worked well on the S60 webkit browser as well and looked and functioned much nicer than the typical mobile websites optimized for the much crappier browsers that were common before that. Opera mini became a thing around the same time, which was an actual browser implemented in j2me that you could run on feature phones.
Performance was not really an issue with this. That was the problem: it worked a bit too well and you didn't need an iphone to use these websites.
Apple fixed it with a native SDK, an appstore, and a by now well established practice of systematically crippling the browser experience on IOS in subtle ways. For a long time you had weird memory restrictions, they never bothered with progressive web apps, and they have very strict policies in the app store that further ensure users and developers focus on building "native" experiences. It would be fairly easy to fix it. But of course there's a rule against third parties doing that too. Safari is the one and only browser that is allowed on the platform. Officially that's because it benefits the user but the obvious actual reason is that Apple does not want people removing obstacles that would allow for a better application experience inside browsers. Flash was banned for the same reasons before the iphone even launched. So were applets and the whole j2me ecosystem that existed at the time.
Google on the other hand shipped chrome on a far more open platform. But they make money from ads rather than hardware sales. Which is why the playstore exists and why anyone earning money with apps and ads prefers using native apps as well. The status quo is these two fairly locked down platforms and a few niche platforms that don't seem to move the needle much when it comes to what people build.
That might change in the next few years as the Apple/Google duopoly slowly heads for inevitable court cases which might introduce more application stores and platforms. When that happens, web first becomes a cheap strategy to target mobile because testing and building apps for each platform is already getting quite expensive with just two of them. Add wasm to the mix and what's native and browser based becomes kind of blurry in any case. Five years is about right in terms of timeline. But you can bet that Apple will drag their heels with all of this.