As far as I can tell, it’s more about what the final output is used for. If it’s to be rendered on a screen, then the convention is that the z-axis is depth (since the x-y plane represent the location on-screen). So game assets and VFX tend to prefer Y-up. If it’s a digital representation of a real object (like in CAD programs), then the more “physical” convention is the useful one.
As far as I can tell, it’s more about what the final output is used for. If it’s to be rendered on a screen, then the convention is that the z-axis is depth (since the x-y plane represent the location on-screen). So game assets and VFX tend to prefer Y-up. If it’s a digital representation of a real object (like in CAD programs), then the more “physical” convention is the useful one.
Narrowly dodging how screen coordinates are Y-down. Even engineers thought that was too silly.