Valve is focused on making a device that works well with their existing game catalog. It's a Steam device first, and it needs to be inexpensive to compete with Quest (which is subsidized by Meta), so they need to prioritize which features get included. I wouldn't be surprised to see a first party AR camera attachment a while after launch. The expansion port seems specifically designed for this, with the inclusion of MIPI CSI lanes for two cameras.

I wonder if this will be a VR trojan horse.

The Steam Deck was wildly popular for a non-Nintendo device. It's got Linux up to 3% of total Steam playtime. If this has a similar draw (play every game on Steam without having to buy a TV), maybe the install base of VR will grow to a point where it's more feasible to make games that support it.

It also makes SteamVR relevant again in a world where Oculus has been eating a lot of the mindshare by releasing affordable headsets and buying the most successful game studios.

I get that there needed to be tradeoffs, I just disagree with this particular one. I could suggest many other ways to save ten bucks in hardware costs. Any other cost saving measure would still allow to play the same games, just with worse performance. But this choice cuts the stock device off from an entire class of apps - in my opinion the best of them all.

I don't think the greyscale camera is mainly a cost concern. I imagine the greyscale camera has better low light and noise performance, which is quite important for tracking.

The big difference seems to be that this headset doesn't have AR cameras at all, but reuse the mapping camera for some light passthrough duty.