This is really concerning from a system point of view, and something I've feared since they started with (NEF) Home Connect in building cloud integrations directly into the miniserver.
Architecturally it means that random external internet servers out of their control are being integrated more closely into the core of a house building control system, and the opportunities for those servers to have random updates that subtly change some behaviour and cause a Loxone crash taking the entire house offline. It's sad if your light switches and fire alarm no longer work just because the washing machine cloud service has pushed an update.
Add to that, as the number of integrations goes up the number of potential interactions between them grows exponentially, and the frequency of Loxone firmware updates will have to accelerate to keep up with random service updates being pushed on a timeline out of their control, which is itself a vector for new bugs getting into the system. So I expect this is only going to get worse over time, and expect some zero-day exploits to pop up too.
I much prefer the v1 architecture of a "local connection only" miniserer and using some separate, more powerful (and non-mission critical) gateway to bridge to cloud services, e.g. Loxberry, iobroker, home assistant, 1home, or whatever. The main drawback I see is Loxone themselves don't have a strong support or recommendation for this architecture, and never will now they're trying to build everything into miniserver v2