That is an interesting reversal. You could also say that the SDK for Spotify is available [1], but HomePod is yet to take advantage of them. The Spotify SDK has been available for many, many years and is integrated into many devices already.
There is a long history of Apple changing the rules on Spotify [2]. Given Apple's focus on promoting their streaming services, I would not be surprised if Apple is intentionally trying to pressure Spotify into a narrow lane as far as integration options. They wouldn't want to cannibalize their own streaming ecosystem.
Apple historically has not shown much interest in integrating direct support for third-party services, even in spaces they aren't competing in. In the few instances it has, these integrations saw little maintenance and were eventually removed. This makes some amount of sense with how rocky third-party APIs can be[1][2]. Generally, it's easier for both ends when service providers write their own integration for a given platform.
As an aside, the timeline given by Spotify is somewhat disingenuous because it doesn't include any of the times that APIs were opened up and is missing significant details about why rejections happened — for instance, early watchOS rejections (as far as I recall) had to do with the lack of a power efficient streaming API in watchOS. Once Apple had a streaming solution that was fit for public consumption (internal/private APIs almost always aren't and cleaning them up takes time), they added it to watchOS.
Don't get me wrong, I think iOS would be better if it opened up a little, but Spotify is no saint and has a tendency to embellish the story.
Even if Apple didn't have a competitive offering to Spotify, it'd be a terrible idea to adopt a third party's standard/SDK that impacted such a core part of the ecosystem and UX as controlling and handing off audio. The reason people buy into ecosystems is so there's a single entity that is weighing and making decisions about what creates the best user experience, which necessarily means having control over the roadmap of standards and interfaces.
Even if we were designing the perfect tech ecosystem that was free of business or competitive constraints, it still wouldn't be a good idea to have an app developer, with their own motivations that may not be aligned to users, dictate SDKs and standards to ecosystems or platforms. It's a weird and unsustainable power dynamic that misaligns incentives. It'd only be viable if in the process the app developer opened the standard and shared control of its future.
As much as Spotify wants to be a platform, what they offer is just a service, and they're not positioned to be controlling such standards. I love Spotify and much prefer it to Apple Music, but what they're doing is choosing not to build features that their customers want on platforms their customers have chosen in order to gain leverage in the regulatory battle over the 30% cut. I respect the desire to have a more level playing field but their other complaints are disingenuous and are not in the best interest of their customers.
Apple has a position as a gatekeeper and is charging a toll, unjustifiably. It's disingenuous to say that other people submitting to Apple's tolls are not working in the long-term best interest of their customers.
Spotify is slow to implement features their customers want (see: Watch, HomePod, TV, Siri), which is not in their customers best interest. Other smaller services like Pandora seem to be able to do this [1], leading one to reasonably conclude Spotify is choosing not to.
This has nothing to do with the 30% cut, except that its likely Spotify is crippling themselves to make their regulatory case (similar to Epic's approach). See sibling comment [2] regarding Spotify's slant on their timeline post, in which they seem like they're entitled to certain things that the platform is not yet capable of (e.g. Watch apps in 2015 when the APIs were very crude).
Fast forward to 2020 when these issues have long since been sorted out and Spotify still lacks an offline Watch app. As a Spotify customer I'd love to be able to listen to music while on a run, but Spotify has decided this core use case of an extremely popular platform is not worth building.
