There's another way that Threads could gain an advantage, and that is by releasing fediverse software that is really great that people want to use, but that the community has no control over and is released with a restrictive license. This software could provide advantages in local resource and network efficiency, instance management, ease of configuration, ease of upgrade, etc, which would be appealing to people who run instances. Once that software has a big enough share, they could begin introducing changes that suit them and not the community. If ease-of-upgrade is in place, and not enough transparency from meta about the included changes, folks could end up unwittingly adopting bad features, perhaps ones that put them in legal peril and jeopardize their ability to continue hosting the instance. If this software doesn't have easy data export features, then migrating the instance to new software could be an insurmountable task.

I think we should all be incredibly critical of any community and systems maintenance challenges in software released by meta, and be diligent about testing migrate-away scenarios. In fact, I would say that if they do release self hostable software, we make sure to port all the good features to FOSS software as quickly as possible. We don't want to end up in a legal bind by adopting source-available software that is not free-as-in-freedom.