My guess is that the engine will start, but the question is whether other powetrain modules will complain about the powerclass. Since you got a manual transmission, might not be an issue. Immo off typically patches SiaRls (immobilizer release function), it should not prevent it from starting even if powerclass is not the same. But you have to try and see for yourself, tough to guess 100% over the internet.

Thank you guys.
So a little update on the topic.

After rearranging of pins in ECU connectors and Immo off job car starts without problem and ECU have communication with ABS/ESP and Parking brake.
However there is SAFE on the dashboard next to the mileage counter, and I have error in 05-Acc/Start Auth. module for Imo data not adapted in ECU, and another two faults in 01-Engine ECU - one for internal flash checksum error, and another one for start blocked by immobilizer.
Nevertheless car starts and you can rev it without problem.

I think this guy that made immo off doesn't do good enough job and that's why I get those errors.
Let's say it is not a big deal if I can get right fueling with 2.0 Hitachi HPFP.

Soon I'll post my notes for rewiring 1.8 wiring loon in order to use 2.0 ECU on those cars.

Regards,
Chris

That one is a little bit tricky.

I don't have scanmatic so I can use pcm flash for bench read, so I gave the ECU to a guy here.
He send me files that he read originally from the ECU, but I can't be sure about this.
I already noticed he is not so aware what he is doing, and not a well orginized.
For example I asked him to change TPROT bytes in flash after removing immo, so my friend can read/write after that with kess.
He was not able to change those two bytes, nevertheless I gave to him address where he can find them.
So I asked to send me immo off flash file, to change them on my own.
It turns out he send me original flash file, so no immo off.

Now I'm thinking that he managed to do some frankenstein writing back 2.0T flash and 1.8T eeprom - I'll check this later.

So that are flash and eeprom that he send me for original files - let's hope they are original and form the same ECU 

Yeah I know it is possible, but I don't have scanmatik hardware to do it on bench.

I do this only for my cars, and trough OBD only.

That's why I should delegate this things to some other peoples that have the hardware and don't know how to use it 

Anyway - if I found a bench hardware right now I'll read the same shit that is inside the ECU right now

Let's hope this guy sent to me real original backup (flash+eeprom) from the 2.0 ECU (the files that I include in rar file above).
 

The Xor method in eeprom does not always work and also does not stop the error messages to the Dash.

The method that can be used to pflash on old med17 ecus needs 2-3 things.

1.Bypass siaRLS (Return 1)
2.Bypass DTC on siaError (Tell function that there was no error)
3.Kill CAN message to dash on a certain ini Function.

There might be an easier method but this is what worked for me on the oldies.On newer Ecus like simos18/19 med17.1.62 or mg1 things are much easier.

As promised - here you can find rewiring that should be made in ECU loom, so you can use 2.0T Med17.1 ECU instead of 1.8T Med17.5.

https://nefariousmotorsports.com/forum/index.php?topic=4996.msg165400#msg165400

At first I forget to move signal from G40 Hall sender - car starts but with longer cranking - it seems ECU manage to deal with missing camshaft reference signal quite well.

It probably does. Most likely it was not done properly.