As I'm reading _Diffusion of Innovations_ (which is super problematic in many ways, making it a very difficult read), it occurs to me that three really crucial factors are inhibiting adoption of consumer cloud services.
(1) Most people who can afford to pay for consumer cloud services are old enough to already have some sort of media solution already: they've got an enormous iTunes library on their desktop or a harddrive or whatever. They are used to curating media; they have a backup system and they are either uninteresting to malicious hackers or have come up with a way to protect themselves. To convince this group of people to switch, you have to make the case that the cloud offers something significantly better AND not significantly worse where significantly is defined as Enough to Justify the Trouble of Switching. Your cloud service had better be fast, convenient, cheap and not prone to horrifying and unexpected failure. Also, your consumer has to believe you will stay in business.
(2) Many people whose media life is large/complex enough as a consumer to want consumer cloud services are going to challenge your ability to scale your service to meet their needs, at a price point they can be happy with. If you cannot satisfy these power users, they will moan to everyone around them, who will be reluctant to bother to research your service because their early adopter friends are dissing you.
(3) Cloud services are inherently difficult to explain. They are basically non-local file systems with multiple access points with security, backups and so forth managed by someone else for a fee. Problems can arise at the connection (if the consumer doesn't understand the data transfer requirements, they could cost themselves a bunch of money on data overages on a phone plan, be frustrated by how long things take, etc.), at access (consumers are prone to forgetting usernames and passwords, which cloud services tend to use, and if you're too lax about security to ensure the customer can access then you run into the next problem), while keeping them secure, maintaining consumer confidence in privacy, etc. Worst of all, however, is maintaining integrity of data while allowing changes to the data through multiple access points. The whole point of iTunes in the iCloud or iTunes Match or whatever is that once your music is there, you can listen to it on your phone, your tablet, your laptop, your AirPlay connected speaker. Apple has trouble maintaining Purchased Library consistency, and Amazon has trouble shipping Purchased Library data -- at least for large libraries. Things get exponentially worse when the customer can change the data and those changes aren't successfully synchronized across devices. Google Drive solves this by not letting you do _anything_ offline (mostly but not completely true); EverNote attempted to let you work offline to a greater degree, and the results for years were unpredictable and occasionally devastating. Both solutions can profoundly interfere with productivity and produce enormous customer frustration.
If you're using EverNote or Google Drive, and you understand the many contributors to delays in accessing your data to view or change, you may or may not be able to figure out where your problem lies and correct it. But most users not only don't understand the many additional components involved in a client-server system vs. a monolithic system -- they aren't ever going to be able to understand them. If adopting, or continuing to use a product or service with a cloud component _requires_ understanding too much, that product or service will limit its market and likely ultimately fail.
It'll be interesting to see how this shakes out. A lot of people never, ever, ever did backups. Some fraction of them ultimately lost data that was of some importance to them. The cloud offers some convenience and security by offering to maintain your data for you, at some cost of accessibility now and going forward. It is _not_ obvious to a capable user in the existing world of desktops, laptops, mini desktops, etc. that many consumer cloud services (uploading your music to the cloud, vs. an external hard drive you maintain; uploading pictures to Flickr, vs. etc.) are a win. But most potential consumers of cloud services are _not_ capable users in the existing world. It remains to be seen whether these services can be presented in a sufficiently simple manner to sell better and sustain the ongoing usage patterns that justify the subscription model.
