I don't know if you'll ever see this, but thought I'd reply.
First, the original Rift headsets were as you describe: lightweight, passing through the PC VR image. However, Meta did not miss out on an opportunity. In what was perhaps the most effective A/B test they could run, they released the Rift S (tethered PCVR) and Quest 1 at effectively the same time. The market feedback was resounding: I believe it was a 10-to-1 preference for a standalone experience vs. tied to a PC. Since they doubled down on standalone (or all-in-one if you prefer), well over 20 million headsets have been sold. In fact, they're so popular that even the fraction that connects to Steam is basically tied for market share with the most popular PC VR headset ever, the Index.
Second, even as a PC VR HMD it was a real stretch to call it a monitor equivalent. It's wildly complicated to create compelling VR images. You need two screens at nearly 2Kx2K resolution each, running at 90 frames/second, sustained. Dip below that and you can induce nausea. Not every PC can do that, so you need careful engineering between the client and HMD, with tricks like time warp, space warp, interleaving, compression, prediction, pose estimation, etc. to take up the slack. Creating sub-millimeter precision of location with six degrees of freedom either requires external base stations (cost, complexity) or inside-out tracking with headset-mounted cameras and a processor running realtime simultaneous location and mapping and image recognition code, which implies a CPU and tech stack to support it. Nowadays people also expect passthrough (with real-time depth correction), hand tracking (AI routines for hand posing), and more. All this is to say that significant code must run on the HMD for a modern gaming headset (Meta's target market), as well as on the PC. And if you're investing that much in a custom software stack, you can't make it up on hardware margin - the cost to build an HMD is just too high. So you have to have an app store tie-in, because Valve sure isn't going to share its Steam profits with you.
Now, certainly there have been (and are) HMDs that tried this approach. HP (G2) and HTC (Vive series) both put out quality products leveraging the Steam ecosystem. Neither are sold in volume today, because the economics of selling a headset just aren't good enough.
Immersed and Big Screen are releasing very lightweight fixed-function HMDs for either work or movie watching that do operate the way you describe. Neither are expected to be high volume devices, and both are more expensive than Quest 3.
In short: VR is much, much harder than you may realize. Meta didn't miss an opportunity, the explicitly chose the market-tested, most popular solution that also has an economic model with some potential future payoff. If you want a "minimal experience akin to an OSD" then look at the Big Screen Beyond ($999, https://www.bigscreenvr.com/) or the Immersed Visor ($1,049, https://www.visor.com/). (Note: compare the price of these hardware-model pass-through devices to the Quest 3 ($499) which also includes a CPU, battery, storage, audio, more RAM).
It's also worth noting that Quest 3 is not online-only. It works fine offline once you've logged in once (people use it on planes, in parks, in the car, etc.). But this particular issue at Meta forcibly logged out users, then the API appeared online while failing all future login attempts. Ironically, users that work offline never noticed the outage because the bug couldn't log them out.