qwe said:
... that'd place us at the center, which is silly. ...
You have the right idea. It would be
quite silly to limit our idea of the U to an observable sphere around us.

Because models which extend beyond that can be tested by what they predict that we
can observe.
Restricting our models to the immediately observable range would make them unnecessarily and unreasonably complex. Nobody does it and there would be no justification for it. So you are right. Silly is the correct word.
But you need to check the definition of the Hubble sphere, and be sure what the different horizons mean.
the observable universe extends to the Hubble sphere right,
Actually not Qwe, the observable extends way way beyond the Hubble sphere. The Hubble radius corresponds to redshift z = 1.4, ( BTW that is a
proper distance of 13.7 Gly. Proper means freeze-frame, it would be what you would measure by conventional means if you could freeze the expansion process at this moment, and depends on estimated values of cosmo parameters).Anyway we see way beyond z = 1.4. We see millions of galaxies with redshifts larger than 1.4, out to more than z = 7! And the Background has redshift about z = 1090!
When we observe the microwave background we are looking at matter which is now at a proper distance of over 45 Gly. If you could freeze expansion it would take 45 billion years for a flash of light to get from here to the matter that emitted the background radiation we are now receiving. Of course that matter used to be a thousand times closer but that is how far it is now.
By definition the Hubble radius is merely the distance that is expanding at rate exactly c. If a galaxy is right at z = 1.4 then (if it is stationary relative to background) the distance to it is increasing at rate c. If it is farther, with larger redshift, then of course the distance is increasing at more than c.
The Hubble law v = Hd is defined in terms of proper distances and the current rate they are increasing.
But anyway, just to be clear, the observable is way bigger than the Hubble.
but there's no evidence that everything would just stop exactly at the line of Earth's observable universe.
That's right! It's an important point. It would be foolish to pretend the universe "stops" at a proper distance of 45 Gly (the distance today of the most distant matter we are observing.) In cosmology we work with a
mathematical model and our view of the universe is
inferred. The standard model certainly does not stop at 45 Gly!

It wouldn't work if it did.
The game in a math science like cosmo is to get the simplest model that fits the data. And nobody I know of uses a bounded model where the U "cuts off" at some observation limit like 45 Gly (proper).
A few (pretty clearly misguided) people might say we OUGHT to only consider the universe what we can see in real time----mathematical inference based on model-fitting should not be allowed, they imply.
Heh heh, but almost NOTHING is being eyeballed real-time. A dragon could have eaten everything out beyond redshift z = 0.1 yesterday and we wouldn't know it for a long time!
Almost everything we can say is based on inference using a model (which has been tested every way we can think using all the data we can scrape up!)
We infer how things are nowadays from light we are getting now that was mostly emitted long ago. Strictly speaking that is mathematical model based inference, and relies on the simplicity of the model. You could always postulate a dragon who has moved things around or eaten a bunch of stuff, so that the real U is different from what we infer, but putting in the dragon would complicate the model.
Same way with putting in a boundary with a cutoff at some distance like 45. It would just make things more complicated and there is no reason to do it.