In order for this to happen, those web apps need to have every capability a native app can have, including access to the local area network, if you want to do "things that native apps do". What if I want to write an HTML5 SSH client so that I don't have to rewrite it twice for iOS/Android? An SMB client? A custom streaming protocol? An HTML Arduino IDE that can flash code over the LAN?
Ideally, a web app could simply have a manifest of permissions, present this list to the user at an appropriate time, and the user decide whether or not to grant them. Opening arbitrary TCP sockets should also be a permission, just like native apps need to add this to their manifest. Native apps already ask for permissions when installed; web apps should be required to do the same, and in return, be granted the same set of capabilities.
The idea that "web apps should be restricted" is at odds with the vision of using web apps to realize a true cross-platform app programming ecosystem.
Now it might happen that this will effectively happen as a result of convergence and standardization of browser extension APIs. It'll be interesting to see what happens there.
That said, the manifest of permissions idea is fundamentally broken in at least two related ways:
1) Most users have no rational basis on which to decide whether to grant the permission or not. In fact, I would argue this is all users, due to issue #2:
2) The permissions are either overbroad, such that pretty much any app that wants to do anything interesting asks for them (this is the "all my apps on want to be able to read my contacts and browsing history" syndrome) or there are so many of them that the list becomes way longer than any user would ever read at install time.
You example of a permission for opening arbitrary TCP sockets is an excellent case in point. As a standalone permission, what fraction of users do you think would understand what is actually being granted? Even if you restrict to computer professionals, I suspect that most would not understand the implications without a somewhat lengthy explanation of what the permission actually allows. And even then, in practice it would be rolled into some other more broad permission that would likewise not indicate the security implications of granting it.
http://robert.ocallahan.org/2011/06/permissions-for-web-appl... has some more in-depth discussion of the issues with this model.
> Native apps already ask for permissions when installed
Yes, and in practice everyone just grants whatever permissions are asked for without really even reading.
That's because the browser and the server have different routing tables. Or to put is more simply, the browser can see the stuff on your LAN, behind your firewall, while the server can't. The point of CORS even for anonymous requests is to prevent web pages from being able to exfiltrate stuff out from behind firewalls, so you don't have random data leaks just because someone in an organization visited some random website. (Yes, you also need them to not download and run random binaries, not have any unpatched exploits in their browser and OS, etc... security is hard.)
Now obviously native apps _can_ exfiltrate stuff out from behind firewalls. Which means that if you really care about your LAN's security you have to assume or enforce that none of the phones connecting to it have any sketchy apps installed...
Note that there is in fact discussion about having a concept of "installing" a web app that you trust sufficiently, which would relax the browser security sandbox such that you get more of the capabilities you're talking about here.