It's a nice idea but it won't gain traction as it's still designed with 2000's era computing in mind. Nobody manages an entire complex system within a specific Linux distribution anymore. There is a vast ecosystem of tooling, systems, protocols, networks, services, that comprise modern distributed systems. All these things exist regardless of the Linux distro.

In fact, I will predict right now the death of Linux as a dominant backend computing platform. It's sort of happened already. Sure, Linux is the kernel that runs the host machines that the cloud runs on. But the services are increasingly serverless/kernel-less, or run on micro-vms running either Linux or another kernel, or run in containers made of a half dozen different base distros. All the Linux I/O is just passed to an outer layer with a universal adapter, so you can mix and match networking, filesystems, logging platforms, policy governance, orchestration, scheduling, etc. I spend nearly all of my time building systems by tuning software and services that have almost nothing to do with Linux.

Linux distros will still be around in one form or another. But long gone are the days where most people define their system or security based on a particular distro. Hell, most of the software used in the cloud today isn't even packaged by distros. Linux has become just a kernel again, the distros just a fancy installer for GNU tools.

They're using a lot of advanced tech and advertising having to "maintain fewer virtual computers", whatever that means. But it also seems their tech is more complex and less compatible with existing components. There's no clear idea of how a developer's supposed to take their code on their laptop, test it, and ship it to production as one immutable image. If it's not as simple as containers, it's not going to replace them.

Also, it's interesting that they seem to allude to immutability, but don't mention the principle by name at all. It seems like whomever is developing this doesn't run large distributed systems.