Yes, the upper arms would have to be capable of telescoping to match the arc of the longer lower links. But the rubber bushings don't allow enough movement to make up for this, they just lessen how much you notice when the upper links bind and jam into the chassis. The binding is still happening, and the mounting locations are still taking the full hit.
So spherical rod ends, even if installed on both ends of upper and lower control arms, are not going to induce or create a new binding situation that wasn't already there with a bone stock rubber bushingsbushing oem set-up.
Any time you have unequal length control arms, you will have bind, as they are competing against each other, trying to travel along different arcs. So if it's all but impossible to eliminate said binding, without entirely re-engineering the suspension and changing the body to accommodate new equal length links, then you might as well eliminate any and all slop in the suspension so that at least it's predictable.
Rubber bushings are in no way consistent. And consistency is predictable, predictability inspires confidence. I will gladly deal with having to reinforce the torque boxes, and a slightly harsher ride for a suspension that reacts the same way every time, all the time. It's going to bind no matter what, so I might as well be able to predict when it's going to happen and how the car handles when it does.
That's just me, though. Personally, I think the answer to all SN95 rear suspension questions is the irs swap. It's not the best irs out there by any means, but it sure beats a not the best solid rear axle suspension.
I think the concern here, that perhaps you might be missing, is this: it's a given that it's going to bind. Okay. And just so we are on the same page, binding here means that the suspension wants to move in an undesirable way, or direction that we don't want (read: undesirable way). Alright so we know this. Now, if you have rubber bush's in the UCA's, the movement that will lead to binding is somewhat absorbed in the deflection of the rubber. If the binding is severe enough, the rubber bush's alone will not stop it. They will only soften it somewhat. If the binding is not as severe, it's possible the rubber bush's will absorb all of the stress.
Okay. Now, if we change the rubber bush's in the UCA to spherical ones, they will not deflect. At all. Which means that when the binding comes, the stress does not get absorbed by any rubber bush's, and therefore whatever those spherical bush's are attached to now get the full stress of the binding. If this happens to be mounts that aren't designed for that kind of stress/angle of application, then you will distort the mount. In other words, you will damage the mount. If the mount is strong enough, but the bush happens to be weak, then the bush will distort. If the bush is strong (solid spherical), and the UCA is the weak point, *it* will distort. The point here is that without something to help absorb the binding, something else (that probably also happens to be expensive) will absorb the stress, and most likely distort/fail/break.
I think what you are missing here ultimately is that in NORMAL/DESIRABLE suspension travel, then yes, spherical bush's are super. However, due to the design of our rear suspension, in addition to NORMAL/DESIRABLE movement we also get ABNORMAL/UNDESIRABLE movement. This latter is what causes the issues with solid bush's.
The solution, obviously, is to remove the items that cause the ABNORMAL/UNDESIRABLE movement, in this case the 4-link. Swap to a torque/arm, panhard, and remove the UCA's, and run spherical bush's to your heart's content.