In short you are correct that they are both providing an ip address to queries for FQDNS within their configurations. However, the WideIP is looking at the health of the hosts that you have in WideIP's pool and has the ability to apply logic where Zonerunner(Bind) could care less and it will return the ip address it has on file. If there is more than one, it will round robin with out any logic.
A Wideip can make decisions on which ip address to return to the clients based on topology, cpu, packet rate, number of hops and various other data it has on hand. This works best when the ip address being returned is a VIP from an F5 LTM, as then the GTM module can pull the status and other data about that VIP via iQuery to allow for a better decision to be made as to which ip address should be returned the client and what is in its WideIP pool.
An example would be say you had a Widip with two ip address in its pool, one address is for the US datacenter LTM VIP and other address is for the Europe datacenter LTM VIP. When a dns request comes in , you could use topology records keying off countries to tell the wideip to only return the US VIP IP for clients requests from the US and the same for Europe based requests to only get the Europe VIP IP. Then if say the US VIP goes down, the Wideip see that status change and would return the Europe VIP IP for US clients as backup untill the US VIP comes back up.
Or another way you could always have all clients get the US VIP ip address, but when it goes down they will get the Europe VIP ip address. All depends on how you configure the WideIP and GTM