Hello Colm,
first of all I would verify that EIGRP neighborship is built on vlan 61 between SW-sideA and R2-SideB.
The second note is that it is possible that in normal conditions R2-SideB uses vlan61 to reach the route and so it is seen as a downstream router:
to be listed in the eigrp topology table as a feasible successor the route advertised by R2-SideB must pass the feasibility test:
metric or R2-SideB < best metric to the route
if R2 uses the path via SW-SideA it is seen as more far from the destination route and so it is not installed in the EIGRP table.
Actually R2-Side B itself once decides that the path via vlan 61 is better applies split horizon and removes its advertisement of the route out vlan 61
(doesn't advertise out an inteface what it learns on the same link)
So what yo see is correct behaviour because the EIGRP metric is proportional to sums of delay and to the inverse of lowest bandwidth in path.
To check just do
sh ip route network
on R2-Sideb and see if the outgoing inteface is vlan61 to SW-SideA
Hope to help
Giuseppe
