[fuzzy2]

I’d like to quickly address the second paragraph. From my experience, iPhones last until their software support runs out, barring any accidents. Furthermore, PWAs need not be offline applications (so TFA overblows the data loss potential ever so slightly) but often work exactly like you want them to.

I use the Outlook PWA. It’s great because I don’t have to let my company manage (part of) my phone.

[aragilar]

Sure, caches make sense, and are where PWAs are a legitimate solution. But PWAs also are sold as a replacement for native apps (hint: they're not, but a whole bunch of native apps should be PWAs, or just static sites...), and the author of TFA does subscribe to that view.

[fuzzy2]

But they can replace a lot of native apps. Calendar, Notes, Weather, Maps (to an extent), Reminders, Calculator, $CellProviderApp, Stocks, Cloud Drive (Apple, Google, Microsoft, …, to an extent), Translator, Mail, various Messengers, Youtube, Amazon, various public transport ticket apps, EV status apps, parcel tracking apps, …

I’m only on home screen 2 of 5 and oh my, it’s almost all of the apps. Granted, some would perform worse in some ways. Some could not realize all features (like automatic photo upload). More integration (like sharing to PWAs) would be needed. But some others? Yeah. Why even develop an app?

[aragilar]

Sure, and a bunch of those are already on the web, though it depends what you're trying to do e.g. mail could just be a view of an existing service (in which case it's just making the existing interface mobile friendly), or it could be an actual client (using IMAP/SMTP), in which case it should be native, rather than having some weird intermediary (which if you want to to do IMAP/SMTP via the web, you need to do). Personally, I'd rather native apps with well defined interfaces that use well defined standards (e.g. IMAP/SMTP/ical), rather than some hacked together web version. But people like reinventing wheels...