More shutdown fixes

[jstnlef]

Description

As it turns out, timing shutdowns is very hard to manage. I ended up putting the "stopping" variable back into the member list to ensure that it doesn't call Cluster().ShutdownAsync() again before it actually gets to the System.ShutdownAsync(). This will ensure that it is only called once if it's initiated by itself. The next thing I did was to change EndpointTerminatedEvent's OnError property to ShouldBlock because I found that even in a graceful termination, the code right now will attempt to reconnect to a node that it "just" terminated because the gossip actor likely has messages in its mailbox where the "Respond" PID includes the node that should be removed. By pausing reconnections, the errors I was seeing around reconnects seem to have disappeared.

Purpose

This pull request is a:

• Bugfix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)

Checklist

• I have added tests that prove my fix is effective or that my feature works
• I have added necessary documentation (if appropriate)