Will Web Applications be Fixed Soon?

I read recently that in-browser apps are extremely favored over native apps (people use their favorite social media app and maybe one other and don't bother with anything else). This makes sense; users are taking the path of least resistance - don't bother installing something unless you have to.
But yet, in-browser applications are so frequently busted. They even seem to get worse as time goes on in some cases. I used to listen to slacker and pandora in the browser. Slacker stopped working some time ago (in any browser), and pandora stopped working recently. I mean not working as in not at all - can't make it through more than one song without hanging the browser. So I started using spotify which seems to default to having you install a client. It works. You can even pause your song and it resumes at the exact place the next time you open the app. The UI is terrible IMO, but it works. Isn't this the goal, to actually work?
Web as a platform for applications still seems like a hack after all this time. I remember getting excited about Silverlight back in the day: yay, we can write an application and it will work like an application! It was a workaround for web as platform. Then Silverlight (and Flash) died.
Now Service Workers / WebAssembly seem to be the next candidate to overcome the web as platform hack. Will they enable applications to actually work in browsers, or will they go the way of flash? Both are starting out as actual standards, so they seem promising.
If they do catch on, won't that render most of the native app work obsolete?  Maybe these technologies can't help whatever broke slacker/pandora?  What other capabilities will remain un-achievable at that point?