I had a lot of similar problems with PETG initially. After trying many things with the profile settings like you seem to be doing, it turned out all I needed to do was use the generic PETG profiles with a “K value” set to 0.030; plus use a release agent on the textured plate (glue stick or similar).
How to do this: In Bambu Studio, on the Device Tab, right hand side, “Ext Spool” box, there is a tiny area that looks like a pen and paper to click. Hit that, you’ll get a dialog box in which you can enter the filament type and K value. For my personal filament and setup, with a 0.4mm extruder the default K value of 0.020 had it not sticking whatsoever - same results you are getting; after calibration I set the K value to 0.030 (for a 0.4mm nozzle) or 0.050 (for a 0.6mm nozzle) and things printed perfectly with PETG.
So perfectly that it fused itself to the textured print surface tearing it up - ignore what the Bambu chart says and use a glue stick (or something similar) barrier on the textured plate when printing PETG!
Hopefully this 0.030 setting will get you close - now in the device tab of Bambu Studio hit the ‘CALI’ button on the bottom right (next to the Unload and Load buttons). All this does is send a pre-canned gcode file to the printer that prints a one layer high calibration pattern with different K values, the idea being you can visually pick the K value that looks the best for your printer / filament / environmental situation. One of the problems with this approach is if your filament setting / K value isn’t close to start with, you effectively can’t print out a recognizable calibration plate since nothing will adhere to the plate if your initial K is too low; or worse yet just produce a huge blob of plastic if your initial K is too high. Note that Bambu Lab pulled this method of calibration from the SoftFever fork of Bambu Studio.
Material management and calibration isn’t flushed out for the P1P, and Bambu Lab has admitted as much in their postings saying they are working on it. At the moment the calibration method described above is clunky, error prone, but it does work and when it is set up properly the printer prints PETG like a champ. If you ever wondered what the LIDAR on the X1C was good for - this is probably it.
Note that the Generic PETG profile has very conservative setting the flow rate at 10mm/s which less than half that for Bambu specific PLA filaments, so it prints a lot slower. I ran a series of prints on my filament to test flow rates of 10, 12, 14, 16, 18, and 20 mm/s; varying the temperature a bit for the higher flow rates. For my case I found 10, 12, and 14 mm/s worked well; anything higher and I was noticing visual artifacts but still acceptable prints. 20 mm/s certainly worked for me, but I doubt I’d make the tradeoff in quality very often to use this setting with PETG; I am usually printing between 10 and 14 mm/s depending on the results desired.
I have two different brands of PETG available (Prusament and Filimant Depot); both seem to work well with the generic PETG profile and a K value as described. Other brands of PETG will probably need their own unique settings, but should be somewhat close to that described above.
So in conclusion if you use some sort of release agent with your textured plate (I’m using the Bambu Lab liquid glue intended for the cold/warm plate stickers) and get through the calibration process, you should have excellent results from this printer in PETG. Hopefully a more seamless calibration solution will be introduced after some more development.