Blockchain-based systems are... not simple, though they do do work. The advantage there is... what? The ability to verify sender/recipient? At the cost of massive power consumption to keep the encryption going, an unnecessary level of distribution for most purposes, and, as best as I can tell, the inability to delete anything that nobody involved cares about anymore.
If anything replaces email, it'll be basically email with security built in (end-to-end encryption and the guarantee that the From address matches the sender). More than that (maybe a little extra, I might have missed something) is overkill. Even that much is unlikely to come up except as patches on a per-provider basis, which might hopefully become omnipresent enough to remove support for the old version.
Email is simple, powerful and everyone understands it. Yes, we need more security, better filtering and all, but not a replacement for it.
If you do this in the European Union, you are shooting yourself in the foot.
Why?
First because data protection laws dictate anyone can oppose against the treatment of his/her data in your systems. Email is considered personal information here. If you put email in a blockchain you cannot effectively remove data from it because it breaks the whole chain of signatures.
Maybe you could do something like cyphering the data before dumping it in the blockchain, and using a smart contract or inserting a special record to make it unavailable (not erased, but made not accessible).
Second, you cannot distribute / share personal data as you want. In some countries -like Spain- you need a signed autorization by the Director of the Data Protection Agency.
It works somewhat like TOR in that it will depend on volunteers to host nodes that will communicate with existing email providers (outlook, gmail, yahoo, etc). The reason we need these nodes is that it's practically impossible to host an email server at a residential location because ISP's block port 25. Also most people don't have the desire or knowledge to maintain an email server.
Users can host their own delivery nodes (which will be plug and play software pre-installed on RPi's). These nodes communicate with the volunteer's host nodes over 80, thereby bypassing the ISP block of port 25.
There will also be an option for people hosting delivery nodes to share their resource with people who can't/won't host their own nodes which will allow people to sign up for email accounts without running their own software (a la gmail).
It gets way deeper than that, but I'll leave it there. You can follow the project at https://github.com/freedatafound/. I've also bought http://freedatafoundation.org but there's currently nothing there.
It is similar to TOR in the sense that it will rely on volunteers to host nodes and/or relays. The goals of TMail and TOR are, however, fundamentally different.
