Their businesses model, as they make clear in their kickstarter page, is the same as others, take 30% rev share of the games (most of the critics seem to have not read that). Which is a multi-billionaire fast growing industry on Android. There's nothing obviously wrong with them selling hardware cheap to make up for it with games.
And their insistence on free games with in app payment only shows they understand this industry much better than most game developers. Which gives them a huge leg over slow incumbent giants like nintendo and MS. Who are still struggling to understand what the hell this freemium thing is for. (tip: if you see a game developer use terms like "DLC" or "DRM", it's a red flag that they're not catching up with our time)
I'm not sure what freemium would be other than "DLC"
The games won't all be free, there just must be at least a demo.
I daresay that they will be able to hire a bunch of Chinese android devs (e. g. the MIUI project) for less than 0.5 mil as opposed to 6 mil.
They have the independent air for now, but I think the chance of them taking funding from somewhere is pretty high in the long run for many of the reasons you state.
Now historically game console manufacturers sold consoles near (or below!) cost and then made up the difference selling licensees as part of the games, made famous by Nintendo early on, the 'custom game cart' that they had to make for you. But these guys want to keep games low cost (even free) so that's a challenge.
Then there is the OS, sure its a port of Android but it isn't Android. It has their own stuff in it, and like all complex software packages Android is full of bugs of varying complexity, and being a bit 'off axis' as they are they will probably wander into less tested code. So they no doubt have at least one and probably two or three full time Android kernel/system engineers. So that's like a .5M$/yr burdened cost, more so if they do point optimizations to improve their user experience.
We've got the Founder(s) (presumably the designer is on contract and they can jettison those ongoing costs once they are done) so say they run a really tight ship, maybe 10 people? So maybe they have a $6M/yr burn rate. How much do they have to sell to cover those costs? What numbers did they promise their suppliers? Lets be generous and say they can sell 100K units the first year. That would require clearing $60/unit to break even. They currently have about 13K backers, so for everyone one they sell on Kickstarter they will need another 10 to sell somewhere else.
Or, perhaps the Kickstarter project is all there is. Once they make the 13 - 15k units (or what ever it ends up to be) for the developers they are done. They walk away. What sort of future does any console have when rounded to the nearest million units there are zero out there ?
So its an interesting thing indeed. Having played with Android on the Pandaboard I can say its not a 'fun' place to be (well if you're trying to get something done, if you just want to hack Android its great fun) Obviously you can't know if it will be a runaway success or a well meaning implosion but either way its enlightening to see this crowd funded business model tried for this class of goods.