Since all of the comments on this page so far seem to be opposed to the integration:
I love Firefox's "Reader View". It's the only way that some pages are readable on my desktop or mobile device, because so many websites try to hijack scrolling, insert modal overlays and ads, or do all sorts of things that make it unbelievably frustrating to just read static text.
On the other hand, Reader View lacks a sync feature. There was one for a few days in Nightly, but it was buggy for the short time it existed, and it was removed.
I was hoping Firefox would improve the sync feature and bring it back eventually, but in all honesty, this is way better. The work that goes into making a seamless, syncing reader view is not trivial[0], and it makes more sense for Mozilla to focus on building a browser than to reinvent the wheel when Pocket already exists and works incredibly well with the same use case.
As for whether this should be "bundled" into the browser vs. an extension: I agree that it would be nicer philosophically if Pocket were a preinstalled extension. On the other hand, Firefox Hello is literally a preinstalled extension with no special integration or privileges (other than being preinstalled), and still some people made the same complaint about it when it launched[1]. So I take that complaint with a grain of salt.
And as for the performance impact of either, I'd have to see some data demonstrating that this actually leads to an appreciable (let alone measurable) increase in memory or CPU usage to be convinced that simply not using it is not an acceptable alternative.
[0] it may look that way, but there are a lot of corner cases
[1] From what I understand, Firefox Hello is simply an extension that leverages WebRTC features already built into the browser to enable video chat (with the assistance of a service provided by Telefonica, which assists in the routing).