I think this way: we haven't announced s0/1 link into EIGRP 10 due to the restriction announcing only transit networks in task 4.1, so 156.1.45.0/24 wouldn't be known outside of R5, what effectively breaks the direction to have full IP reachability at the end of the lab.