I don't think it's the entire premise of this Slack organization, and the definitions are definitely unclear, especially with marketing nonsense around it. Platform Engineering has a much broader scope than Backstage or even DevOps/DevSecOps.
A platform is about abstracting a value stream away from the specifics of infrastructure and turning it into something that works for developers and end users in a uniform and reliable way. It can include some or all of the tooling needed to bring workloads to production and keep it there, with observability, management, lifecycles for various components, etc.
An internal developer portal, if it exists, should live on that platform. That platform has internal and external components that service developer needs and the needs of end users, both.
A good example of this is a multi-IaaS, multi-region solution based on kubernetes that abstracts infrastructure away from developers, that provides a place for the Internal Developer Portal to live. That's at least the technical side of a platform. My point is that if it's internal, it's not really meeting the definition of a platform and if it's serving end users, it's no longer Internal or Developer.
It's an important semantic distinction, because words mean things. I've seen a lot of folks assert that k8s on its own, or Backstage on its own, or even a stack of hand-managed Jenkins VMs in an IaaS are all Platforms, when none of those things encompass the entirety of a capital P Platform.
Cloud Foundry is a good example of a fairly fully fledged platform technology.
Also, for the record, this naming thing is not really the big problem, just an indicator of other challenges.
The really hard part isn't the tech, but the way it's operationalized as a product in service to developer, end user, and other stakeholder needs. Platform teams need to engage those folks and ask them what they need, prioritize what they see as important, measure, adjust, rinse, repeat.
Both Platforms and Portals need to be operated as products.