I was reading the article Some Tips on Making a FETching Discrete Amplifier by George Alexandrov and Nathan Carter and was wondering about the stability of this amplifier. They make the statement "The circuit shown in Figure 2 can achieve comparable noise at unity gain, without the need for compensation. The speed is primarily determined by the operational amplifier. " Is this because the speed of the JFET, LSK389A, is much greater than that of the ADA4897? For classic compound amplifier design for stability it is best practice to have the dominate pole of the design in the front-end of the amplifier. So either the LSK389A is fast enough to not effect the phase and gain margin of the ADA4897 appreciably or it does limit this amplifier and this statement isn't really accurate.