BG-TH Network Purpose?

I have recently been learning a bit about simulating Basil Ganglia-Thalamus networks using the built in networks, but I’m having trouble understanding what the purpose of these networks are. As far as my understanding goes, you take your utility values for each action and feed it into the network and then get a WTA output for the selected action. What I don’t understand is that it seems that the network just seems to select an action based on which one has the highest utility value, so then what is the point of using a simulated neural network when you could write a function to just select the action with the highest utility value? What is the point of the BG-TH network?