I had been working on, during my time away from Steem the architecture and protocol specs for a one-datatype one-database, non-blockchain structured distributed database and application framework that takes distributed systems to what I think is the next level - automatic geolocation sharding, fractal epidemic back-propagation and regular digest reposts to allow potentially, sub 100ms clearance times, in a provisional form, for distributed databases like Steem.

You can see the beginnings of the documentation and initial parts of the protocol specification here:

https://gitlab.com/dawn-network/nexus

The reason why I am responding to this post in this way, is because what I am proposing could implement, even better than just a wallet with 50 different distributed database 'wallets' in it, a mechanism whereby this inter chain commmunication can be 'centralised' in this system, and allow the binding of all blockchains together into a cohesive whole.

This would include exchanging on chain, monetising the delivery of uploaded media from users for small nodes providing local replication, it would facilitate further experiments in the architecture of distributed database, and a critical element also, alongside allowing the aggregation of a user's many crypto accounts into a single account database from which they can manage them, is to enable 'in band' instant and long form messaging (email and instant messaging) which is bound to these accounts and can be also third party verified through challenges, enabling seculity while greatly increasing utility of cryptocurrencies and appchains like Steem, indeed LBRY as well, and potentially, a foundation with which to implement the long promised Fabric of Steem.

I would thus like to suggest that this may represent a means towards continuing the adoption and proliferation of use cases for Steem, beyond the current, somewhat difficult to expand framework of Steem's system.