Can Ethereum be the backbone of the new internet?

With an exciting recent announcements regarding progress on Plasma, sharding, and Casper, 2018 seems like it could be a big year for Ethereum. The online community has at times discussed Ethereum as leading the push towards Ethereum 3.0 - is this a realistic vision, and how much closer towards this vision would the completion of these three projects take us?

Sharding will aim to preserve the security and decentralization of Eth while increasing performance, by splitting up ethereum addresses into separate shards, then having assigning only a small subset of nodes to process transactions and store state for each shard. If effective, this would increase performance by a constant factor based on the # of shards, let's say by ~100x.

Plasma will allow for some transactions to be processed by "child chains" which rely on the main chain for security. The child chains can themselves have child chains, and can process transactions like the main Ethereum chain, or even ahve their own sets of rules. In the event of an attack (e.g. 51% attack) on a child chain, the child chain can be shut down and valid transactions summarized and re-written onto the main chain, allowing good users a safe exit. In theory, Plasma would allow for potentially "infinite" scaling, as most calculations could be moved to child chains.

Well these two projects are impressive as of themselves, one gap that remains is the need for decentralized, secure, and easily accessible data storage. Storage costs on the blockchain are currently very expensive and impractical for the larger amounts of data which would be needed by some modern applications. While Plasma could be a partial solution in allowing data storage in child chains, to be a universal computer, Ethereum will need a good off-chain storage solutions as well.