It seems there is a misunderstanding of how the system works.
You create multiple profiles, each representing a virtual coordinate space that is applied based on specific qualifiers. The first qualifier defines the virtual space size. The engine selects the closest qualifier according to Defold’s dynamic profile selection. Once the engine selects a profile, it applies all the overrides from the GUI file associated with that profile.
You can specify a separate profile for each screen size, but it’s not mandatory. The engine operates using a virtual profile space, which is then mapped onto the real space using the rules specified in the nodes. For example, if the selected profile includes a zoom on the background, it will adjust according to the screen size, etc.
So, in your example
You have one virtual coordinate space set to 1920x1080, and you design your interface according to this profile. This profile will be applied if one of its qualifiers fits better than all other available qualifiers. The engine picks the most suitable profile based on the defined qualifiers, ensuring that the interface is adjusted to match the best possible virtual space for the given screen size.