So let us say it is voted on and the answer is keep the mailing lists. What are the next steps to fixing the mail system which is held together by duct-tape and bailing wire? 

It is a seriously large task to get done with everything from someone getting all the packages back into Fedora, architecting how email should flow, getting a documented system and tooling, then dumping the current system, fixing the schema changes, importing the system into a new mail system, and doing that 6 or 7 times to get it actually working. It is a set of tasks needing senior programmers, some DBA time, and senior system administration. It will take about 3 to 6 months of dedicated work to get done. If done on volunteer time, I would put it at 12 months. After that it needs a regular redo and cleanup of code or we will be back where we are currently very fast.

Where we are currently is a system running on old dead code, can't be reinstalled, with a possibly slightly corrupted disk partition (though I think I fixed all of the items) and needing to be coddled regularly to keep going. It has hundreds of thousands of 'held' dead spam messages, similar number of bounces for dead accounts (mailman3 did not implement auto-unsubscribe til after the version we are running) and other oddities. 

Or the infrastructure team can work on keeping the build system running which is also fairly touchy with daily restarts, random reboots, etc. That is also more than a full-time job for both the volunteers and dedicated people. 

Look I am not wanting to move to discourse, but voting is not going to fix anything.