Hi Kilian
I have attached a screenshot of both versions I think you had in mind; it does not matter, if it is exactly what you have realized in your amp, because it shows the principle.
I take off the signal right before the second tube at its grid.
As you can see, "vergleich02" shows an arrangement well known of the non-mv Marshalls treble input - values do vary (I know) but this does not matter, as stated before.
"vergleich01" shows an arrangement very similar to most mv-Marshalls - values do vary, too.
The resistors 900k/100k do simulate a 1M log gain pot which is turned up half.
What can be seen at first is the difference in frequency response because in the second arrangement the 470k resistor does act as a partially (frequency dependent) bypassed grid stopper.
At second you will notice a big difference in amplification at 100Hz and still at around 1kHz.
Regards, Stone