I would stick with mid.ssh.use_snc=true unless you have a reason not to. At the very least, it has a separate error message for connection and auth failures.

When use_snc is true, what error message do you get?

If it's an auth failure, check credentials. Does your version have the "test credential" feature? That would be a good place to start.

If it's a connection failure check that you can get to connect to port 22 on the target from the mid server.

If all else fails, you can set mid.ssh.debug=true, recreate the failure, and look in the agent log for a tsunami of detail on what's going on.