The transformer does not have a feedback winding. Maybe I don't understand your post.
The transformer is center tapped for the circuits power supply.
First stage is a charge amp cascode (15pf is always in circuit) pads are more caps added to reduce gain
I don't like how charge amps sound I don't understand why people like the fet 47 circuit as well
Second stage is a follower with a CC in the source leg with switchable flat and two high pass setting
The CC setting at T3 will change the voltage to the first stage so for best results.
I would look for voltage readings from a real one
The CC resistor to adjust would be R8 that will help set the circuit up.
If you change R9 it will mess up the biasing in the HPF settings
If you change the first stage fet
You don't want to change R6 because it will change the output resistance of the first stage.
R5 would be a good place to start. I would make a temp fixed and variable to find a good value and then install a resistor(s) of the same value.
Note this is combination bias fixed and source biasing
If you look at other AKG circuits 220uHs are used for RF filtering however the RF caps can be a lower value .01uf
Have you looked at this thread?
https://groupdiy.com/index.php?topic=74723.msg946423#msg946423
If you substitute a fet look at the Vgs and IDSS curves for a hint for what you might need to change