I'd settle for one well-thought-out UI concept, built from the ground up, targeting specific device categories, and running on a modern, legacy-free foundation.

I'd settle for one well-thought-out UI concept, built from the ground up, targeting specific device categories, and running on a modern, legacy-free foundation. That would not be suitable for every user or use case and shouldn't replace a more legacy-oriented Windows, in my opinion, but it'd serve a large contingent of users and OEMs designing target devices better than what feels like a hodgepodge of hardware support and UI philosophies, accumulated over multiple decades. Unfortunately, while modern-day MSFT does extremely well financially[0] with their current strategy, their ability to execute such a major OS project from the ground-up and stick with it to the finish line is doubtful given historical examples[1]. Given that they are doing so well financially, it's fair if they feel that such investments aren't worth their efforts, though I view it as potentially shortsighted, considering the value of continued ownership of a large percentage of the platform market. Even if they become a transformer-model-focused company on top of being business over end-user focused, having future business users familiar with your system and retaining some control over the software that your transformer-model implementations run on is sensible. If one wants what is described in the article and shown in the attached video, there are solutions out there to get very close to that UI. Combine that with LTSC and I feel that would scratch that particular itch quite well. [0] https://www.windowscentral.com/microsoft/microsoft-is-getting-close-to-overtaking-apple-as-the-most-valuable-us-company https://www.windowscentral.com/microsoft/microsoft-is-gettin... [1] https://www.windowscentral.com/surface-duo-andromeda-windows-core-os https://www.windowscentral.com/surface-duo-andromeda-windows...