It's not pointless at all. It makes the Web platform more complicated to implement, which, as the company that owns the leading browser, is in Google's interest. The more complicated the platform becomes, the harder and more expensive it will be for any potential competitors to ship a product that can challenge Chrome.

More importantly, as it seems to require an implementation of native UI and not just rendering, it will harm browsers that are currently Chromium based.

I meant pointless to web tech and consumers. Obviously Google has their own reasons.

The spec goes into detail of why existing tech (iframes) aren't enough:

https://github.com/WICG/portals/blob/master/explainer.md#faq...

https://github.com/WICG/portals/blob/master/explainer.md#alt...

To be clear, this FAQ isn't actually part of the spec (which I'm counting as the document here: https://wicg.github.io/portals/).

Link prefetch/preload puts the loading under the control of the browser. The user can't see whether it loads or not, so the browser can make intelligent decisions about when to load it or whether to even load it at all.

Portals are user-visible, so the browser can't defer the load.

> Link prefetch/preload puts the loading under the control of the browser

This is a good thing. Many are on metered connections and have data caps.

Clunkier? To me, this new tag proposal screams AMP, which Google wants to remain fast.

Speed is not the only measure of clunkiness.