The problem is that the question of what is a "browser" is an easy question for humans to answer but doesn't line up to software.
The short story is that the permission model would have to be changed; they have to add an Android permission for sharing browsing data that apps would have to declare and users would have to opt into. And of course that increases the risk surface area; We know from past examples that malicious apps sneak through (or non-malicious apps previously signed off on end up purchased by someone else and become malicious) and users really don't understand security.
The state is extractable via Google take out and the safest solution might honestly be to not support this directly on Android and instead require Mozilla to figure out how to import a takeout archive via the cloud.
The short story is that the permission model would have to be changed; they have to add an Android permission for sharing browsing data that apps would have to declare and users would have to opt into. And of course that increases the risk surface area; We know from past examples that malicious apps sneak through (or non-malicious apps previously signed off on end up purchased by someone else and become malicious) and users really don't understand security.
The state is extractable via Google take out and the safest solution might honestly be to not support this directly on Android and instead require Mozilla to figure out how to import a takeout archive via the cloud.