We actually test using that Lagom gradient and have commented on it in the review. Refer to observations from Lagom in the contrast and brightness section. We did not see anything there or anywhere else in our testing that was at all unusual for an 8-bit VA panel. It is incorrect to assume that any banding visible would be down to the fact it is ‘6-bit + FRC’ when in fact 8-bit source material outputted correctly using 8-bits (FRC or otherwise) could still and often does still have a degree of banding. Users are also not used to seeing the level of detail that can be brought out in games etc. on a VA panel, especially in dark scenes, where inherent banding in the source material is visible due to the monitor displaying things correctly rather than it being masked by IPS glow or less obvious due to weak contrast.
The panel data sheet has no mention of dithering, AOC have confirmed multiple times that it is a true 8-bit panel (including their product manager who is in charge of the panels used for various products) and I have no reason to think (as above) from my testing that this is not the case. Furthermore, HP and manufacturers more broadly are prone to errors in their specifications, so they are not a reliable source.
P.S. They use variants of the same panel with different panel border thicknesses (hence obvious aesthetic differences). The curvature is actually exactly the same (even though AOC chooses to specify it differently to HP) as this is related to the panel itself and does not change between monitors or in this case very similar variants of the same panel. The panel curvature is probably some way between 1800R and 2000R in reality, for both models. Or perhaps varies slightly between units. AOC America actually quote 1800 – 2000R.