I've seen a lot of people downplay the Steam Deck when that was announced. Even worse when the Portal came out. No surprise it's happening again with a possible Xbox handheld.
This has potential that is getting handwaved away: An open system that natively plays your Xbox games, easily plays PC/console Gamepass, can access other launchers libraries, and cloud streaming all out of the box without tinkering is a very attractive package. All while doing something their current home console doesn't - appeal to the crowd beyond their Xbox fan base. If they manage to make a streamline windows OS for this it'll be fucking fantastic.
Okay but here's the rub: if it lets you access Steam, GOG etc. the way many of you think it will....how cheap do you think this handheld will cost? The moment Microsoft releases an Xbox anything that allows access to Steam, there goes a massive chunk of their B2P & MTX revenue out the window. There also goes a chunk of their Game Pass revenue out the window because no one will expect them to provide access to Steam or EGS on an Xbox handheld, but still keep online play gated behind Game Pass (for non-F2P titles).
So you either get a handheld well below Series S specs (maybe even below Steam Deck specs) at a "cheap" $299, $349 or $399 (because they still need decent profit margins), or you get a more performant handheld that's going to set you back $599, maybe even up to $699, because again, margins. MS wouldn't be able to take that risk of allowing other storefront access without a way to guarantee profits for themselves, so that's where the hardware gets more pricy. And considering such a handheld also risks having owners leave Game Pass behind (again in favor for one of those other storefronts), that would also be factored into any potential price. That price would then dictate the volume these get produced in, which could then again also have a direct influence back on the price if the volume ends up being too low, due to a small TAM.
That's why, and I'm gonna keep doubling down on this, it probably makes the most sense for them to stick with Xbox OS for the next line of hardware. Worry about making/integrating a fully 1:1 Windows gaming UI until the actual endgame. Use Xbox OS, get support for Windows apps & programs through running extended Windows code in some translation layer of Xbox OS. Do that in order to natively enable functioning of alternative storefronts (without needing Valve, Epic etc. to build custom apps themselves) and a Windows Store with whitelisted 1P & 3P Windows apps/programs.
Monetize access of those storefronts through a Game Pass subscription tier (Ultimate obviously, and maybe some new tier mainly focused on those storefronts). Use that Game Pass leverage to soft-subsidize on the hardware (so slightly lower margins upfront on hardware, larger overall margins with Game Pass subscription tied in). Then structure Game Pass tiers so per-month pricing is cheaper if you lock-in for a year or two, vs. paying month-to-month (SIE already do this with PS+ and most sub services in general do this).
That should probably be Microsoft's play for the immediate next generation of hardware; if they allow OEMs to build their own variant systems then that's just a bonus (personally, a SEGA-themed system with its own custom UI frontend with a SEGA console-like shell, new SEGA controller, KB/M and some retro SEGA emulated games thrown in would be really cool). Whenever they make a gaming frontend/UI for Windows that's 1:1 with the console experience (or better), but can also facilitate past & present Xbox, Windows, and DOS games, will probably be when they no longer need to even make gaming hardware outside of peripherals anymore. That'd also coincide with no real vehicle of their own to "push" Game Pass, though at that point they'd probably want Game Pass on everything including Sony & Nintendo systems (also means they'd have to go fully 3P/multiplat, which they're probably going to start doing more wholly starting next year).