The Bubba oscillator section says that 0.5V is fed into the non-inverting amp in order to bias the op-amps to 2.5V (mid range). At first glance it looks correct, non inverting amp, gain of 1+R_F/R_G=5, and 5*0.5V =2.5V. But R_G is not connected to GND; it's connected to the output of an op-amp biased at 2.5V. If the op-amps are biased at 2.5V, then both ends of the R_G R_F voltage divider are at 2.5V (at DC), so the Vinn terminal of the 4x op-amp is also at 2.5V. This can only happen if Vinp is at 2.5V, not 0.5V.
As the circuit is drawn in the app note, I believe the op-amps are biased at 0.5V.
I suspect that this was supposed to be 0.5*Vsupply (which is 2.5V since the supply is 5V).
