Not directly related to this, but question to gurus is, that when table reading goes over max x or y axis, does the ecu read the max x or y from the table at that case?

Next question to gurus: dsg map tuner said, that he can't rise the clutch pressure (more than certain amount?) and therefore the ecu has to give the real torque to to the dsg to achieve good clutch pressures. Is this ecu send torque possible to alter by some variable or is only chance to change the range of the load?

I think, that I try my self first. Have to look, that is it possible to do a repeater to the can-bus after engine, by which would repeat all messages but change the ecu send torque value by boost pressure or something.

Before I decided to attack the ECU, I did exactly this. I made a CAN interface that parsed messages on the fly and "corrected" the values sent. I spent hours and hours down this path. I gained total control of the CAN bus yet still did not achieve what you wanted to; increase the DSG clamping pressure.

I managed to verify that the DSG was receiving my altered values yet it still was not happy. The relationship between torque demand, actual torque, and many other torques, is critical and MUST be done by the ECU. It is far more complex than you think right now.

By all means go down the same path I did and I will help you with as much as I know, but really you are better off spending time on the ECU instead.

They really should need no explanation; all are needed for a properly functioning torque model and accurate fueling.

The idea that you can just slap FI on a NA motor and not worry about the ECU is pure folly.... I have no idea why this persists. I suppose partly out of ignorance and partly out of wishful thinking.

These are the kind of questions that require a million of other questions answered. Like "why does the plane fly?"

In short:
You're going with single turbo and a single downpipe - so why install 2 lambdas? It can be done this way too, sure, but why?
>100 load patch is needed to make motronic "accept" boost pressure.
"Boost control patch" is needed to regulate part throttle conditions and not have the "on-off" car.

Precisely, they trim away from each other, so you end up with -25% on one bank 25% on other and then limp mode.
100% load - the ecu won't let you request more load, because it is NA and 100% means atmospheric.
Boost control is really twofold - the ECU needs to know when it can go WOT and when to use the throttle. It needs to know pre-tb pressure or at least an approximation to control the throttle correctly, and it needs to control boost somehow.

But really - read the FR and search for SY_TURBO, it will answer a lot of your questions.

I am not entirely sure what you are after. But if you want to experiment, here is my breakdown of the CAN command:

outgoing.data.bytes[0] // throttle pedal status (idle / depressed)
outgoing.data.bytes[1] // Engine torque 0 to 410NM (as reported in VCDS)
outgoing.data.bytes[2] // rpm LSB
outgoing.data.bytes[3] // rpm MSB

outgoing.data.bytes[4] // unknown
outgoing.data.bytes[5] // accelerator pedal % 0-100
outgoing.data.bytes[6] // engine torque (loss) 0-416NM
outgoing.data.bytes[7] // unknown

Also bear in mind the DSG is more complex than set and forget. It replies to the ECU. For example it tells the engine to reduce torque so that it can change gear. You can prove this because it will refuse to upshift if you hold the torque at 100%, for example.

There are various behaviors like this, that you sort of get a feel for during experimentation.

Despite my analysis, I failed to achieve what I wanted to and have now chosen to investigate the situation from the ECU side.