next test on cooling is finished. And I have again included print speed for the steep overhangs where I had the feeling, that much slower speeds could make the difference.
Small story on the last test: Actually, I misunderstood the ranges for overhang speed. the overhang percentage says, how much of a loop is hanging in the air, so the angle depends on the layer height. e.g. for 0,2mm layer height and 0,42mm line width, 50% means an overhang of 0,21mm which is roughly 45°. 100% overhang means the complete outer wall is printed over air, and the first inner wall is exactly over the outer wall one layer below. that would calculate to 63,4°. For smaller layer height, the same overhang % is a much steeper angle. e.g. with 0,1mm layer height and everything else the same, 100% equals to ~ 76°, but should print equally well. Very smart implementation indeed.
And an interesting lesson for me: I was really amazed that 63° overhang angle on 0,2mm layers means that the outer wall is not at all supported by the layer below. Yet, this is printable with reasonable results and even very good results if you print very slow as this test showed. Quite amazing in my opinion.
Back on topic:
I printed 55° and 65° overhangs, which is accidently very close to 75% and 100% in my setting. I printed at 20, 40 and 60 mm/s and each with 10%,20%,30% and 40% part fan.
I only did the test for the IEMAI PETG up to now. Maybe I do it for Sunlu tomorrow.
As expected from the last test, 55° overhangs printed flawless at any speed and any fan setting.
The 65° overhangs were perfect at 20 mm/s almost independently from the fan setting. Only at 10% there was a very slight curl at the very end, which didn’t appear at 20% fan. The faster prints were not so great throughout. More fan made them better but never good.
So I think, I will go with 20mm/s for the range 75%…100% and 0 (full speed) for the other ranges. I’m still undecided if I want to increase fan speed in the material profile from 10% to 20%. It still seems reasonably low to not decrease layerbonding. Anyway, the IEMAI PETG seems to be very indifferent about many of the printing parameters like temperature, speed, cooling,… Maybe that is due to the high speed tag?
Please note: This is the first time I spent so much time on optimising print settings for an individual filament. Well possible that many other PETG brands behave equally well if you put that much love into them 
Up to now, the Sunlu seems to be just a tad behind at an even lower price point. And both put the BBL PETG to shame. I have to see if I have samples of some other brands at home and might just try them with the final settings found here.
Somehow that works fine for any of my filaments.
But probably, the model was rather stupid. And a benchy at every speed 300, only outer walls 250 turned out very nice. So I’m regaining trust again.