Currently, we'd skip broadcasting announcements if we don't have any connected peers to broadcast to.

That is great and we'll retry every 30 seconds, however, it might result in the initial broadcast after startup being always skipped as we spawn the peer connection and broadcast tasks back-to-back, likely resulting in the first iteration of the latter always skipping due to the peers not having been connected yet.

In order to fix this, we should allow the connection task to the broadcast task so that the latter only starts once we're established our initial peer connections.