As these devices are increasingly sold to people who have never plugged them into a PC running iTunes, the storage problem is exacerbated: without being able to dump everything to a (cheap) external drive, users agonize over which memories to delete and many simply balk at the entire process. And this isn't just a problem with updates: people struggle with this during regular use.
So what we have is really two problems: no space to update and no way to simply/effectively/cheaply archive device data. So what is needed are two solutions:
1. Apple needs to reserve storage space for updates. [2]
2. Mobile devices, in general, need a PC-free way to bulk-archive data. Cloud services are great and all, but people are far more likely to walk into a Best Buy and pick up an external drive, than they are to find/vet/pay for Flickr. And even if they're willing to pay for the service, most are not willing to pay the hosting fees for the quantity of digital memories they're creating. And if you can't save everything, you're just back at the same problem of asking people to explicitly cast memories into the void, and they just won't do it. Not happily. Not willingly. And they'll resent every instance where they feel forced.
We're well past the point where the value in this arrangement was obvious.
Why iOS devices, in particular, can't talk directly to Time Capsules (for device backup and a local media archive) is beyond me. It's particularly curious that Apple is ignoring the situation, given that they don't seek to profit from being a cloud provider, and would stand to make a decent chunk of change from putting a Time Capsule (or three) into every iOS home. They could even leverage the AppleTV as yet-another-viewer of this home-archived media and sell a few more of those.
I know it's popular to assume this is all inevitably going "to the cloud". But there's simply too much data being collected. People marvel at how much data is being uploaded to Youtube/Facebook/Instagram/etc. But for every picture or video uploaded to those services, how much data did that one user capture that they didn't upload? And then consider how many users are capturing similar amounts of data but never upload anything.
The growth rate of data capture is far exceeding the growth rate of bandwidth and the affordability of "enough" cloud storage. It's going to be years and years until cloud storage is "there". There's a huge opportunity in the interim.
Beyond iOS, I don't know how every storage vendor doesn't have their own android solution to the problem. Solving this problem can sell an awful lot of hard drives. That you're explicitly selling the safe archive of precious memories is going to mean you can sell a higher number of RAID solutions as well. And the growth rate of captured data means you've got a huge market for selling additional/newer storage devices to your installed base.
[1] Desktop users don't fill multi-terabyte hard drives as often, only inasmuch as they don't connect high quality data capture devices. Hand a person a DSLR or a GoPro and they'll fill every drive they buy. Mobile devices fill more-quickly, more-regularly, largely because they all have high quality cameras built in.
[2] The updates themselves probably should be smaller at this point, but you'd always need room for a full restore anyway. So shrinking the update sizes is really only a bandwidth concern and can't lead to a reduction in the reserved space.