16:9, tiny arrow keys and missing page navigation keys are deal breakers.
That said, 42.1 dB under maximum load of the measured softwares is very good for 16.8mm thickness, 6900HX (75W PL1, 90W PL2), 3070TI Laptop (100W). Even for a 45W CPU and 80W GPU, this would be good (39 dB would be very good for that). 42.1 dB for the specifications of this notebook are so astonishingly good that I can't believe it.
Either 6900HX, this notebook's cooling hardware and its fan drivers all are so great improvements that 42.1 dB have indeed become possible despite the thin chassis and not particularly modest TDPs - or measurement of maximum noise must have been flawed.
There are enough CPU-only tests so that noise meansurement under CPU-only load must be right. The most unfortunately, GPU tests are very insufficient: they focus on 3D but fail to measure machine learning, productivity, creativity and synthetics. Therefore, noise meansurement under GPU load is only known for 3D applications but remains unknown for almost all other applications. For verifying or refuting the astonishing 42.1 dB under maximum load, such tests remain mandatory and essential. Not to mention torturing both GPU and CPU simultaneously.
Conclusion: either we have a cooling wonder or much more complete noise measurements must reveal the truth.