Just going to give my 2 cents on the track-ip thing, this is typically used to track an upstream next-hop beyond ur next-hop gateway to have a more robust way to detect an internet failure. For example, you are using a cable modem for internet access, and you are concerned your cable modem will not lose power, but your cable connection to the ISP's next hop router will be down but not your own next-hop (the onsite modem), thus no route failover will occur and you will be in an outage. IP tracking is attractive for this scenario.
If you're using a route based tunnel on both sides, the tunnel inherently does IP tracking for all the routes on the tunnel, because the tunnel will only be established if it can reach the remote peer IP terminating the tunnel. This is an implied IP tracking of the peer gateway, which the routes depend on to be in the table at all.
On the leased line you are directly connected and configuring tracking for a directly connected IP address is a bit wonky.
If you run OSPF on the tunnel interface and give them both a cost of 10, and on your leased line as well and give both ends a cost of 1, you will be able to add more networks to your locations by simply adding them into OSPF at their respective sides.