Applying Middlebrook method for open loop gain measurement

[bonzer]

Hello everyone! Could anyone help me please by taking a look to the way I applied Middlebrook method in this case. Here we have an amplifier with two bjt, the second one forms a closed loop by its own (it's kinda like a common emitter) and the first one is just there with and closes the whole circuit in another loop. I want to open the inner loop at the upper branch and apply middlebrook there as this way I open both loops (even the outer). As said by Rosenstark I should apply it directly tho the active part of the circuit (dependent generator of the BJT).

Should I apply it to both transistors to study the open loop gain and therefore the stability?

If I apply it to the other (first transistor) I get the open loop gain of the outer loop with the second loop closed letting it to behave like a single amplifying block. Is this needed to or it's enough to study the inner loop?