I discovered strange behavior from either the ADA4940 model or the LTspice itself. The second assumption seems more accurate.
As you can see in the screenshots above, splitting the gain resistor changes the gain settings of the circuit.
Is this a know issue? And if so, what is the solution/workaround?