The funny thing is that the code I currently use doesn't use TPS for antilag, just for no lift shift and for launch speed adjustment.
E.g. If you are not completely stopped and have your foot on the gas all the way down, you cannot adjust the launch speed. I would post a video but I don't wanna brag  

Personally I only think it happens when you shift too slow.  What the acceptable duration is, I don't know.

Tried few different technics first trying very hard fast shifting then settled for less aggressive change which I found better for me and feel of the car/power


Glad not just my car getting this issue can't wait now for code now ,makes car I think so much quicker though the gears, like to time before/ after on 1/4 mile run

When I analysed NLS code line by line comparing values with the dump 8D0907551M_AlNls_v2.zip I found following in NoLiftShift part of function:
AccPedalThreshold is defined in .bin with value E6 FF which won't work if map defined as byte and you leave it default or will try to change the byte. There should be word.
I guess it was supposed to be E6 00 = 230 * 0.392157 = 90 grad
The part of code which compares threshold reads word from 017E08H address:
0004A : C2 F9 08 7E       MOVBZ    R9,7E08H   ;AccPedalThreshold, 81H:7E08H (017E08H in bin), value = E6 FF???
0004E : 40 49             CMP      R4,R9
Therefore IMHO there should be word E6 00 and AccPedalThreshold map should be defined as word.

P.S: already sorted out, MOVBZ: Moves and zero extends the contents of the source byte specified
by op2 to the word location specified by the destination operand
op1. The contents of the moved data is examined, and the
condition codes are updated accordingly.

Are you using stock coils? Or did you convert to later 2.0 TFSI coils?

You need code that doesn't change dwell time...

0 dwell time is what makes the coils explode.

I have done this before on other ECU's.

Guess it's time to write this for ME7 too...