I don't think you're making an honest comparison if you're doing it with used parts. You can get lots of things cheaper used.

Definitely not less than half the price. The build I see linked in this thread comes in around $950 with as close to equivalent specs as possible. The machines people on YouTube were building when the specs were announced, before the RAM and SSD prices spiked, were coming in a little under $700 at the time.

How recently have you checked? Word I've gotten from elsewhere is that this is pretty competitive with what it would cost to build it yourself.

The one build I see in this thread is within $100 of the Steam Machine with comparable specs, without the form factor. I'd call that competitive. But if you disagree, they're also releasing the ISO for SteamOS, as they're more confident in broad compatibility, so if you were ever in the market for such a machine, you can save the $100.

The handful of companies that can afford to spend $100M+? Sure. There are only so many of those, and plenty of them go bankrupt after spending that much.

There's an ocean of hardware requirements between the upcoming PS6 and what it takes to run pixel graphics games. Many customers are still happy on PS4 level hardware, and third party titles like Madden still got PS4 versions until just last year.

As a realist, I don't see any way cloud gaming services are an option that customers en masse will be willing to pay what the providers have to charge to make a profit. Stadia was not that long ago, and Google couldn't make it work under what had to be a softball toss for that business model.

AI did a number to gaming, but truthfully, gaming technology was probably about to stand still anyway. Barely any studios can afford to make a game that's so technologically advanced that it pushes our current hardware to its limits.

I think that risk is still reduced because the parent wouldn't have their own Steam account to reserve one. For at least months, anyone buying one of these is probably going to be someone very intentionally doing so.

I don't think there's a danger of that person accidentally buying this thing, as you can only buy it from Valve and not Walmart. From a casually overheard conversation on the subway last week, I can tell you that someone who is seemingly "the average NBA 2K player" is pissed off at the upcharge for PS+ just to make his game function in multiplayer, and that guy is aware of the Steam Deck.

I suspect the DRM those streaming services rely on will start to budge on Linux in the near future, as it gains market share. In the meantime though, I've mostly been using Jellyfin lately, which ironically works on Steam Machine but not the other consoles. I've still got my PS4 from 10+ years ago for the regular streaming services, when I still have a subscription to any of them.

The DRM on Linux browsers that all of these services rely on doesn't let you stream any better than 720p.

They wouldn't be licensing it to Valve for this device; it would be the DRM vendor updating their software for web browsers on what had been a niche operating system and is becoming less and less niche of late.

By the way, the Steam version also doesn't have split-screen for some reason. That's how far we've regressed.

UPDATE: They clarified stating the previous requirements were in error, and both players on PS5 will need user accounts on the console but not PS+ unless you're also playing online co-op. I mean, this is still stupid and needlessly complicated, but it's better than what was originally stated.

But why use one DNS service over another? Sorry if you've covered this already and it's just not clicking yet or something.

This would be how a business commits suicide, not to mention upset their retail partners that sell their hardware.

I don't think there's a mechanism for them to tell if you've got a license for a disc without putting CD keys in the box. OP only needs them to honor ownership long enough to resell the copy.

None of us can predict the future, so we don't know which games will end up on GOG one day, but your plan seems solid enough based on what we know now and what you value.

Yeah, I've only got a handful of services I want to run. It's possible that the bug bites me and I want to go deeper into this stuff, but for the here and now, I'm only eyeing 8-10 things I want to host, and they ought to work across a mini PC and a NAS.

Why do I need that? From my perspective, it seems like it would be more useful if I had far more services that I intended to run than what I'm actually planning for.