As those two things are at totally different places in their product lifecycles and will likely have entirelly different uptakes, possibly different populations and quite different usage patterns (something technically extremelly important in high-performance settings because of caching), in terms of Software Architecture it doesn't make sense to have one on top of the other.

The only thing that makes sense from the point of view of software systems design is Single Sign-On (i.e. linking accounts).

So either their head techies are not really all that senior (frankly, I wouldn't be surprised, as judging by their library design and completelly messy product lifecycles, even Google seems to have few or no genuine Technical Architects) or the decision has nothing to do with technology.

The issue is that large tech companies rarely restrain themselves when there’s a big pile of money to be made. The odds are high that Meta will exploit this whether it was an accident or not.